ncut-pytorch / app.py
huzey's picture
add directed ncut (test)
5dac5bc
raw
history blame
170 kB
# Author: Huzheng Yang
# %%
import copy
from datetime import datetime
import pickle
from functools import partial
from io import BytesIO
import json
import os
import uuid
import zipfile
import multiprocessing as mp
from einops import rearrange
from matplotlib import pyplot as plt
import matplotlib
USE_HUGGINGFACE_ZEROGPU = os.getenv("USE_HUGGINGFACE_ZEROGPU", "False").lower() in ["true", "1", "yes"]
DOWNLOAD_ALL_MODELS_DATASETS = os.getenv("DOWNLOAD_ALL_MODELS_DATASETS", "False").lower() in ["true", "1", "yes"]
if USE_HUGGINGFACE_ZEROGPU: # huggingface ZeroGPU, dynamic GPU allocation
try:
import spaces
except:
USE_HUGGINGFACE_ZEROGPU = False
if USE_HUGGINGFACE_ZEROGPU:
BATCH_SIZE = 1
else: # run on local machine
BATCH_SIZE = 1
import gradio as gr
import torch
import torch.nn.functional as F
from PIL import Image
import numpy as np
import time
import threading
from ncut_pytorch.backbone import extract_features, load_model
from ncut_pytorch.backbone import MODEL_DICT, LAYER_DICT, RES_DICT
from ncut_pytorch import NCUT
from ncut_pytorch import eigenvector_to_rgb, rotate_rgb_cube
DATASETS = {
'Common': [
('mrm8488/ImageNet1K-val', 1000),
('UCSC-VLAA/Recap-COCO-30K', None),
('nateraw/pascal-voc-2012', None),
('johnowhitaker/imagenette2-320', 10),
('Multimodal-Fatima/CUB_train', 200),
('saragag/FlBirds', 7),
('microsoft/cats_vs_dogs', None),
('Robotkid2696/food_classification', 20),
],
'Ego': [
('EgoThink/EgoThink', None),
],
'Face': [
('nielsr/CelebA-faces', None),
('huggan/anime-faces', None),
],
'Pose': [
('sayakpaul/poses-controlnet-dataset', None),
('razdab/sign_pose_M', None),
('Marqo/deepfashion-multimodal', None),
('Fiacre/small-animal-poses-controlnet-dataset', None),
('junjuice0/vtuber-tachi-e', None),
],
'Hand': [
('trashsock/hands-images', 8),
('dduka/guitar-chords-v3', None),
],
'Satellite': [
('arakesh/deepglobe-2448x2448', None),
('tanganke/eurosat', 10),
('wangyi111/EuroSAT-SAR', None),
('efoley/sar_tile_512', None),
],
'Medical': [
('Mahadih534/Chest_CT-Scan_images-Dataset', None),
('TrainingDataPro/chest-x-rays', None),
('hongrui/mimic_chest_xray_v_1', None),
('sartajbhuvaji/Brain-Tumor-Classification', 4),
('Falah/Alzheimer_MRI', 4),
('Leonardo6/path-vqa', None),
('Itsunori/path-vqa_jap', None),
('ruby-jrl/isic-2024-2', None),
('VRJBro/lung_cancer_dataset', 5),
('keremberke/blood-cell-object-detection', None)
],
'Miscs': [
('yashvoladoddi37/kanjienglish', None),
('Borismile/Anime-dataset', None),
('jainr3/diffusiondb-pixelart', None),
('jlbaker361/dcgan-eval-creative_gan_256_256', None),
('Francesco/csgo-videogame', None),
('Francesco/apex-videogame', None),
('huggan/pokemon', None),
('huggan/few-shot-universe', None),
('huggan/flowers-102-categories', None),
('huggan/inat_butterflies_top10k', None),
]
}
CENTER_CROP_DATASETS = ["razdab/sign_pose_M"]
from datasets import load_dataset
def download_all_datasets():
for cat in DATASETS.keys():
for tup in DATASETS[cat]:
name = tup[0]
print(f"Downloading {name}")
try:
load_dataset(name, trust_remote_code=True)
except Exception as e:
print(f"Error downloading {name}: {e}")
def compute_ncut(
features,
num_eig=100,
num_sample_ncut=10000,
affinity_focal_gamma=0.3,
knn_ncut=10,
knn_tsne=10,
embedding_method="UMAP",
embedding_metric='euclidean',
num_sample_tsne=300,
perplexity=150,
n_neighbors=150,
min_dist=0.1,
sampling_method="QuickFPS",
metric="cosine",
indirect_connection=True,
make_orthogonal=False,
progess_start=0.4,
):
progress = gr.Progress()
logging_str = ""
num_nodes = np.prod(features.shape[:-1])
if num_nodes / 2 < num_eig:
# raise gr.Error("Number of eigenvectors should be less than half the number of nodes.")
gr.Warning("Number of eigenvectors should be less than half the number of nodes.\n" f"Setting num_eig to {num_nodes // 2 - 1}.")
num_eig = num_nodes // 2 - 1
logging_str += f"Number of eigenvectors should be less than half the number of nodes.\n" f"Setting num_eig to {num_nodes // 2 - 1}.\n"
start = time.time()
progress(progess_start+0.0, desc="NCut")
eigvecs, eigvals = NCUT(
num_eig=num_eig,
num_sample=num_sample_ncut,
device="cuda" if torch.cuda.is_available() else "cpu",
affinity_focal_gamma=affinity_focal_gamma,
knn=knn_ncut,
sample_method=sampling_method,
distance=metric,
normalize_features=False,
indirect_connection=indirect_connection,
make_orthogonal=make_orthogonal,
).fit_transform(features.reshape(-1, features.shape[-1]))
# print(f"NCUT time: {time.time() - start:.2f}s")
logging_str += f"NCUT time: {time.time() - start:.2f}s\n"
start = time.time()
progress(progess_start+0.01, desc="spectral-tSNE")
_, rgb = eigenvector_to_rgb(
eigvecs,
method=embedding_method,
metric=embedding_metric,
num_sample=num_sample_tsne,
perplexity=perplexity,
n_neighbors=n_neighbors,
min_distance=min_dist,
knn=knn_tsne,
device="cuda" if torch.cuda.is_available() else "cpu",
)
logging_str += f"{embedding_method} time: {time.time() - start:.2f}s\n"
rgb = rgb.reshape(features.shape[:-1] + (3,))
return rgb, logging_str, eigvecs
def compute_ncut_directed(
features_1,
features_2,
num_eig=100,
num_sample_ncut=10000,
affinity_focal_gamma=0.3,
knn_ncut=10,
knn_tsne=10,
embedding_method="UMAP",
embedding_metric='euclidean',
num_sample_tsne=300,
perplexity=150,
n_neighbors=150,
min_dist=0.1,
sampling_method="QuickFPS",
metric="cosine",
indirect_connection=False,
make_orthogonal=False,
make_symmetric=False,
progess_start=0.4,
):
print("Using directed_ncut")
print("features_1.shape", features_1.shape)
print("features_2.shape", features_2.shape)
from directed_ncut import nystrom_ncut
progress = gr.Progress()
logging_str = ""
num_nodes = np.prod(features_1.shape[:-2])
if num_nodes / 2 < num_eig:
# raise gr.Error("Number of eigenvectors should be less than half the number of nodes.")
gr.Warning("Number of eigenvectors should be less than half the number of nodes.\n" f"Setting num_eig to {num_nodes // 2 - 1}.")
num_eig = num_nodes // 2 - 1
logging_str += f"Number of eigenvectors should be less than half the number of nodes.\n" f"Setting num_eig to {num_nodes // 2 - 1}.\n"
start = time.time()
progress(progess_start+0.0, desc="NCut")
n_features = features_1.shape[-2]
_features_1 = rearrange(features_1, "b h w d c -> (b h w) (d c)")
_features_2 = rearrange(features_2, "b h w d c -> (b h w) (d c)")
eigvecs, eigvals, _ = nystrom_ncut(
_features_1,
features_B=_features_2,
num_eig=num_eig,
num_sample=num_sample_ncut,
device="cuda" if torch.cuda.is_available() else "cpu",
affinity_focal_gamma=affinity_focal_gamma,
knn=knn_ncut,
sample_method=sampling_method,
distance=metric,
normalize_features=False,
indirect_connection=indirect_connection,
make_orthogonal=make_orthogonal,
make_symmetric=make_symmetric,
n_features=n_features,
)
# print(f"NCUT time: {time.time() - start:.2f}s")
logging_str += f"NCUT time: {time.time() - start:.2f}s\n"
start = time.time()
progress(progess_start+0.01, desc="spectral-tSNE")
_, rgb = eigenvector_to_rgb(
eigvecs,
method=embedding_method,
metric=embedding_metric,
num_sample=num_sample_tsne,
perplexity=perplexity,
n_neighbors=n_neighbors,
min_distance=min_dist,
knn=knn_tsne,
device="cuda" if torch.cuda.is_available() else "cpu",
)
logging_str += f"{embedding_method} time: {time.time() - start:.2f}s\n"
rgb = rgb.reshape(features_1.shape[:3] + (3,))
return rgb, logging_str, eigvecs
def dont_use_too_much_green(image_rgb):
# make sure the foval 40% of the image is red leading
x1, x2 = int(image_rgb.shape[1] * 0.3), int(image_rgb.shape[1] * 0.7)
y1, y2 = int(image_rgb.shape[2] * 0.3), int(image_rgb.shape[2] * 0.7)
sum_values = image_rgb[:, x1:x2, y1:y2].mean((0, 1, 2))
sorted_indices = sum_values.argsort(descending=True)
image_rgb = image_rgb[:, :, :, sorted_indices]
return image_rgb
def to_pil_images(images, target_size=512, resize=True):
size = images[0].shape[1]
multiplier = target_size // size
res = int(size * multiplier)
pil_images = [
Image.fromarray((image * 255).cpu().numpy().astype(np.uint8))
for image in images
]
if resize:
pil_images = [
image.resize((res, res), Image.Resampling.NEAREST)
for image in pil_images
]
return pil_images
def pil_images_to_video(images, output_path, fps=5):
# from pil images to numpy
images = [np.array(image) for image in images]
# print("Saving video to", output_path)
import cv2
fourcc = cv2.VideoWriter_fourcc(*'mp4v')
height, width, _ = images[0].shape
out = cv2.VideoWriter(output_path, fourcc, fps, (width, height))
for image in images:
out.write(cv2.cvtColor(image, cv2.COLOR_RGB2BGR))
out.release()
return output_path
# save up to 100 videos in disk
class VideoCache:
def __init__(self, max_videos=100):
self.max_videos = max_videos
self.videos = {}
def add_video(self, video_path):
if len(self.videos) >= self.max_videos:
pop_path = self.videos.popitem()[0]
try:
os.remove(pop_path)
except:
pass
self.videos[video_path] = video_path
def get_video(self, video_path):
return self.videos.get(video_path, None)
video_cache = VideoCache()
def get_random_path(length=10):
import random
import string
name = ''.join(random.choices(string.ascii_lowercase + string.digits, k=length))
path = f'/tmp/{name}.mp4'
return path
default_images = ['./images/image_0.jpg', './images/image_1.jpg', './images/image_2.jpg', './images/image_3.jpg', './images/guitar_ego.jpg', './images/image_5.jpg']
default_outputs = ['./images/image-1.webp', './images/image-2.webp', './images/image-3.webp', './images/image-4.webp', './images/image-5.webp']
# default_outputs_independent = ['./images/image-6.webp', './images/image-7.webp', './images/image-8.webp', './images/image-9.webp', './images/image-10.webp']
default_outputs_independent = []
downscaled_images = ['./images/image_0_small.jpg', './images/image_1_small.jpg', './images/image_2_small.jpg', './images/image_3_small.jpg', './images/image_5_small.jpg']
downscaled_outputs = default_outputs
example_items = downscaled_images[:3] + downscaled_outputs[:3]
def run_alignedthreemodelattnnodes(images, model, batch_size=16):
use_cuda = torch.cuda.is_available()
device = torch.device("cuda" if use_cuda else "cpu")
if use_cuda:
model = model.to(device)
chunked_idxs = torch.split(torch.arange(images.shape[0]), batch_size)
outputs = []
for idxs in chunked_idxs:
inp = images[idxs]
if use_cuda:
inp = inp.to(device)
out = model(inp)
# normalize before save
out = F.normalize(out, dim=-1)
outputs.append(out.cpu().float())
outputs = torch.cat(outputs, dim=0)
return outputs
def _reds_colormap(image):
# normed_data = image / image.max() # Normalize to [0, 1]
normed_data = image
colormap = matplotlib.colormaps['inferno'] # Get the Reds colormap
colored_image = colormap(normed_data) # Apply colormap
return (colored_image[..., :3] * 255).astype(np.uint8) # Convert to RGB
# heatmap images
def apply_reds_colormap(images, size):
# for i_image in range(images.shape[0]):
# images[i_image] -= images[i_image].min()
# images[i_image] /= images[i_image].max()
# normed_data = [_reds_colormap(images[i]) for i in range(images.shape[0])]
# normed_data = np.stack(normed_data)
normed_data = _reds_colormap(images)
normed_data = torch.tensor(normed_data).float()
normed_data = rearrange(normed_data, "b h w c -> b c h w")
normed_data = torch.nn.functional.interpolate(normed_data, size=size, mode="nearest")
normed_data = rearrange(normed_data, "b c h w -> b h w c")
normed_data = normed_data.cpu().numpy().astype(np.uint8)
return normed_data
# Blend heatmap with the original image
def blend_image_with_heatmap(image, heatmap, opacity1=0.5, opacity2=0.5):
blended = (1 - opacity1) * image + opacity2 * heatmap
return blended.astype(np.uint8)
def segment_fg_bg(images):
images = F.interpolate(images, (224, 224), mode="bilinear")
# model = load_alignedthreemodel()
model = load_model("CLIP(ViT-B-16/openai)")
from ncut_pytorch.backbone import resample_position_embeddings
pos_embed = model.model.visual.positional_embedding
pos_embed = resample_position_embeddings(pos_embed, 14, 14)
model.model.visual.positional_embedding = torch.nn.Parameter(pos_embed)
batch_size = 4
chunk_idxs = torch.split(torch.arange(images.shape[0]), batch_size)
device = 'cuda' if torch.cuda.is_available() else 'cpu'
model.to(device)
means = torch.tensor([0.485, 0.456, 0.406]).view(1, 3, 1, 1).to(device)
stds = torch.tensor([0.229, 0.224, 0.225]).view(1, 3, 1, 1).to(device)
fg_acts, bg_acts = [], []
for chunk_idx in chunk_idxs:
with torch.no_grad():
input_images = images[chunk_idx].to(device)
# transform the input images
input_images = (input_images - means) / stds
# output = model(input_images)[:, 5]
output = model(input_images)['attn'][6]
fg_act = output[:, 6, 6].mean(0)
bg_act = output[:, 0, 0].mean(0)
fg_acts.append(fg_act)
bg_acts.append(bg_act)
fg_act = torch.stack(fg_acts, dim=0).mean(0)
bg_act = torch.stack(bg_acts, dim=0).mean(0)
fg_act = F.normalize(fg_act, dim=-1)
bg_act = F.normalize(bg_act, dim=-1)
# ref_image = default_images[0]
# image = Image.open(ref_image).convert("RGB").resize((224, 224), Image.Resampling.BILINEAR)
# image = torch.tensor(np.array(image)).permute(2, 0, 1).float().to(device)
# image = (image / 255.0 - means) / stds
# output = model(image)['attn'][6][0]
# # print(output.shape)
# # bg on the center
# fg_act = output[5, 5]
# # bg on the bottom left
# bg_act = output[0, 0]
# fg_act = F.normalize(fg_act, dim=-1)
# bg_act = F.normalize(bg_act, dim=-1)
# print(images.mean(), images.std())
fg_act, bg_act = fg_act.to(device), bg_act.to(device)
chunk_idxs = torch.split(torch.arange(images.shape[0]), batch_size)
heatmap_fgs, heatmap_bgs = [], []
for chunk_idx in chunk_idxs:
with torch.no_grad():
input_images = images[chunk_idx].to(device)
# transform the input images
input_images = (input_images - means) / stds
# output = model(input_images)[:, 5]
output = model(input_images)['attn'][6]
output = F.normalize(output, dim=-1)
heatmap_fg = output @ fg_act[:, None]
heatmap_bg = output @ bg_act[:, None]
heatmap_fgs.append(heatmap_fg.cpu())
heatmap_bgs.append(heatmap_bg.cpu())
heatmap_fg = torch.cat(heatmap_fgs, dim=0)
heatmap_bg = torch.cat(heatmap_bgs, dim=0)
return heatmap_fg, heatmap_bg
def make_cluster_plot(eigvecs, images, h=64, w=64, progess_start=0.6, advanced=False, clusters=50, eig_idx=None, title='cluster'):
if clusters == 0:
return [], []
progress = gr.Progress()
progress(progess_start, desc="Finding Clusters by FPS")
device = 'cuda' if torch.cuda.is_available() else 'cpu'
eigvecs = eigvecs.to(device)
from ncut_pytorch.ncut_pytorch import farthest_point_sampling
magnitude = torch.norm(eigvecs, dim=-1)
# gr.Info("Finding Clusters by FPS, no magnitude filtering")
top_p_idx = torch.arange(eigvecs.shape[0])
if eig_idx is not None:
top_p_idx = eig_idx
# gr.Info("Finding Clusters by FPS, with magnitude filtering")
# p = 0.8
# top_p_idx = magnitude.argsort(descending=True)[:int(p * magnitude.shape[0])]
ret_magnitude = magnitude.reshape(-1, h, w)
num_samples = 300
if num_samples > top_p_idx.shape[0]:
num_samples = top_p_idx.shape[0]
fps_idx = farthest_point_sampling(eigvecs[top_p_idx], num_samples)
fps_idx = top_p_idx[fps_idx]
# fps round 2 on the heatmap
left = eigvecs[fps_idx, :].clone()
right = eigvecs.clone()
left = F.normalize(left, dim=-1)
right = F.normalize(right, dim=-1)
heatmap = left @ right.T
heatmap = F.normalize(heatmap, dim=-1)
num_samples = clusters + 20
if num_samples > fps_idx.shape[0]:
num_samples = fps_idx.shape[0]
r2_fps_idx = farthest_point_sampling(heatmap, num_samples)
fps_idx = fps_idx[r2_fps_idx]
# downsample to 256x256
images = F.interpolate(images, (256, 256), mode="bilinear")
images = images.cpu().numpy()
images = images.transpose(0, 2, 3, 1)
images = images * 255
images = images.astype(np.uint8)
# sort the fps_idx by the mean of the heatmap
fps_heatmaps = {}
sort_values = []
top3_image_idx = {}
top10_image_idx = {}
for _, idx in enumerate(fps_idx):
heatmap = F.cosine_similarity(eigvecs, eigvecs[idx][None], dim=-1)
# def top_percentile(tensor, p=0.8, max_size=10000):
# tensor = tensor.clone().flatten()
# if tensor.shape[0] > max_size:
# tensor = tensor[torch.randperm(tensor.shape[0])[:max_size]]
# return tensor.quantile(p)
# top_p = top_percentile(heatmap, p=0.5)
top_p = 0.8
heatmap = heatmap.reshape(-1, h, w)
mask = (heatmap > top_p).float()
# take top 3 masks only
mask_sort_values = mask.mean((1, 2))
_sort_value2 = (heatmap > 0.1).float().mean((1, 2)) * 0.1
mask_sort_values += _sort_value2
mask_sort_idx = torch.argsort(mask_sort_values, descending=True)
mask = mask[mask_sort_idx[:3]]
sort_values.append(mask.mean().item())
# fps_heatmaps[idx.item()] = heatmap.cpu()
fps_heatmaps[idx.item()] = heatmap[mask_sort_idx[:6]].cpu()
top3_image_idx[idx.item()] = mask_sort_idx[:3]
top10_image_idx[idx.item()] = mask_sort_idx[:6]
# do the sorting
_sort_idx = torch.tensor(sort_values).argsort(descending=True)
fps_idx = fps_idx[_sort_idx]
# reverse the fps_idx
# fps_idx = fps_idx.flip(0)
# discard the big clusters
# gr.Info("Discarding the biggest 10 clusters")
# fps_idx = fps_idx[10:]
# gr.Info("Not discarding the biggest 10 clusters")
# gr.Info("Discarding the smallest 30 out of 80 sampled clusters")
if not advanced:
# shuffle the fps_idx
fps_idx = fps_idx[torch.randperm(fps_idx.shape[0])]
def plot_cluster_images(fps_idx_chunk, chunk_idx):
fig, axs = plt.subplots(3, 5, figsize=(15, 9)) if not advanced else plt.subplots(6, 5, figsize=(15, 18))
for ax in axs.flatten():
ax.axis("off")
for j, idx in enumerate(fps_idx_chunk):
heatmap = fps_heatmaps[idx.item()]
size = (images.shape[1], images.shape[2])
heatmap = apply_reds_colormap(heatmap, size)
image_idxs = top3_image_idx[idx.item()] if not advanced else top10_image_idx[idx.item()]
for i, image_idx in enumerate(image_idxs):
_heatmap = blend_image_with_heatmap(images[image_idx], heatmap[i])
axs[i, j].imshow(_heatmap)
if i == 0:
axs[i, j].set_title(f"{title} {chunk_idx * 5 + j + 1}", fontsize=24)
plt.tight_layout(h_pad=0.5, w_pad=0.3)
filename = f"{datetime.now():%Y%m%d%H%M%S%f}_{uuid.uuid4().hex}"
tmp_path = f"/tmp/{filename}.png"
plt.savefig(tmp_path, bbox_inches='tight', dpi=72)
img = Image.open(tmp_path).convert("RGB")
os.remove(tmp_path)
plt.close()
return img
fig_images = []
num_plots = clusters // 5
plot_step_float = (1.0 - progess_start) / num_plots
fps_idx_chunks = [fps_idx[i*5:(i+1)*5] for i in range(num_plots)]
# with mp.Pool(processes=mp.cpu_count()) as pool:
# results = [pool.apply_async(plot_cluster_images, args=(chunk, i)) for i, chunk in enumerate(fps_idx_chunks)]
# for i, result in enumerate(results):
# progress(progess_start + i * plot_step_float, desc=f"Plotted {title}")
# fig_images.append(result.get())
for i, chunk in enumerate(fps_idx_chunks):
progress(progess_start + i * plot_step_float, desc=f"Plotted {title}")
fig_images.append(plot_cluster_images(chunk, i))
return fig_images, ret_magnitude
def make_cluster_plot_advanced(eigvecs, images, h=64, w=64):
heatmap_fg, heatmap_bg = segment_fg_bg(images.clone())
heatmap_bg = rearrange(heatmap_bg, 'b h w c -> b c h w')
heatmap_fg = rearrange(heatmap_fg, 'b h w c -> b c h w')
heatmap_fg = F.interpolate(heatmap_fg, (h, w), mode="bilinear")
heatmap_bg = F.interpolate(heatmap_bg, (h, w), mode="bilinear")
heatmap_fg = heatmap_fg.flatten()
heatmap_bg = heatmap_bg.flatten()
fg_minus_bg = heatmap_fg - heatmap_bg
fg_mask = fg_minus_bg > fg_minus_bg.quantile(0.8)
bg_mask = fg_minus_bg < fg_minus_bg.quantile(0.2)
# fg_mask = heatmap_fg > heatmap_fg.quantile(0.8)
# bg_mask = heatmap_bg > heatmap_bg.quantile(0.8)
other_mask = ~(fg_mask | bg_mask)
fg_idx = torch.arange(heatmap_fg.shape[0])[fg_mask]
bg_idx = torch.arange(heatmap_bg.shape[0])[bg_mask]
other_idx = torch.arange(heatmap_fg.shape[0])[other_mask]
fg_images, _ = make_cluster_plot(eigvecs, images, h=h, w=w, advanced=True, clusters=100, eig_idx=fg_idx, title="fg")
bg_images, _ = make_cluster_plot(eigvecs, images, h=h, w=w, advanced=True, clusters=20, eig_idx=bg_idx, title="bg")
other_images, _ = make_cluster_plot(eigvecs, images, h=h, w=w, advanced=True, clusters=0, eig_idx=other_idx, title="other")
cluster_images = fg_images + bg_images + other_images
magitude = torch.norm(eigvecs, dim=-1)
magitude = magitude.reshape(-1, h, w)
# magitude = fg_minus_bg.reshape(-1, h, w) #TODO
return cluster_images, magitude
def ncut_run(
model,
images,
model_name="DiNO(dino_vitb8_448)",
layer=10,
num_eig=100,
node_type="block",
affinity_focal_gamma=0.5,
num_sample_ncut=10000,
knn_ncut=10,
embedding_method="tsne_3d",
embedding_metric='euclidean',
num_sample_tsne=1000,
knn_tsne=10,
perplexity=500,
n_neighbors=500,
min_dist=0.1,
sampling_method="QuickFPS",
ncut_metric="cosine",
indirect_connection=True,
make_orthogonal=False,
old_school_ncut=False,
recursion=False,
recursion_l2_n_eigs=50,
recursion_l3_n_eigs=20,
recursion_metric="euclidean",
recursion_l1_gamma=0.5,
recursion_l2_gamma=0.5,
recursion_l3_gamma=0.5,
video_output=False,
is_lisa=False,
lisa_prompt1="",
lisa_prompt2="",
lisa_prompt3="",
plot_clusters=False,
alignedcut_eig_norm_plot=False,
**kwargs,
):
advanced = kwargs.get("advanced", False)
directed = kwargs.get("directed", False)
progress = gr.Progress()
progress(0.2, desc="Feature Extraction")
logging_str = ""
if "AlignedThreeModelAttnNodes" == model_name:
# dirty patch for the alignedcut paper
resolution = (224, 224)
else:
resolution = RES_DICT[model_name]
logging_str += f"Resolution: {resolution}\n"
if perplexity >= num_sample_tsne or n_neighbors >= num_sample_tsne:
# raise gr.Error("Perplexity must be less than the number of samples for t-SNE.")
gr.Warning("Perplexity/n_neighbors must be less than the number of samples.\n" f"Setting Perplexity to {num_sample_tsne-1}.")
logging_str += f"Perplexity/n_neighbors must be less than the number of samples.\n" f"Setting Perplexity to {num_sample_tsne-1}.\n"
perplexity = num_sample_tsne - 1
n_neighbors = num_sample_tsne - 1
if torch.cuda.is_available():
torch.cuda.empty_cache()
node_type = node_type.split(":")[0].strip()
start = time.time()
if "AlignedThreeModelAttnNodes" == model_name:
# dirty patch for the alignedcut paper
features = run_alignedthreemodelattnnodes(images, model, batch_size=BATCH_SIZE)
elif is_lisa == True:
# dirty patch for the LISA model
features = []
with torch.no_grad():
model = model.cuda()
images = images.cuda()
lisa_prompts = [lisa_prompt1, lisa_prompt2, lisa_prompt3]
for prompt in lisa_prompts:
import bleach
prompt = bleach.clean(prompt)
prompt = prompt.strip()
# print(prompt)
# # copy the sting to a new string
# copy_s = copy.copy(prompt)
feature = model(images, input_str=prompt)[node_type][0]
feature = F.normalize(feature, dim=-1)
features.append(feature.cpu().float())
features = torch.stack(features)
else:
features = extract_features(
images, model, node_type=node_type, layer=layer-1, batch_size=BATCH_SIZE
)
if directed:
node_type2 = kwargs.get("node_type2", None)
features_B = extract_features(
images, model, node_type=node_type2, layer=layer-1, batch_size=BATCH_SIZE
)
# print(f"Feature extraction time (gpu): {time.time() - start:.2f}s")
logging_str += f"Backbone time: {time.time() - start:.2f}s\n"
del model
progress(0.4, desc="NCut")
if recursion:
rgbs = []
all_eigvecs = []
recursion_gammas = [recursion_l1_gamma, recursion_l2_gamma, recursion_l3_gamma]
inp = features
progress_start = 0.4
for i, n_eigs in enumerate([num_eig, recursion_l2_n_eigs, recursion_l3_n_eigs]):
logging_str += f"Recursion #{i+1}\n"
progress_start += + 0.1 * i
rgb, _logging_str, eigvecs = compute_ncut(
inp,
num_eig=n_eigs,
num_sample_ncut=num_sample_ncut,
affinity_focal_gamma=recursion_gammas[i],
knn_ncut=knn_ncut,
knn_tsne=knn_tsne,
num_sample_tsne=num_sample_tsne,
embedding_method=embedding_method,
embedding_metric=embedding_metric,
perplexity=perplexity,
n_neighbors=n_neighbors,
min_dist=min_dist,
sampling_method=sampling_method,
metric=ncut_metric if i == 0 else recursion_metric,
indirect_connection=indirect_connection,
make_orthogonal=make_orthogonal,
progess_start=progress_start,
)
logging_str += _logging_str
all_eigvecs.append(eigvecs.cpu().clone())
if "AlignedThreeModelAttnNodes" == model_name:
# dirty patch for the alignedcut paper
start = time.time()
progress(progress_start + 0.09, desc=f"Plotting Recursion {i+1}")
pil_images = []
for i_image in range(rgb.shape[0]):
_im = plot_one_image_36_grid(images[i_image], rgb[i_image])
pil_images.append(_im)
rgbs.append(pil_images)
logging_str += f"plot time: {time.time() - start:.2f}s\n"
else:
rgb = dont_use_too_much_green(rgb)
rgbs.append(to_pil_images(rgb))
inp = eigvecs.reshape(*features.shape[:-1], -1)
if recursion_metric == "cosine":
inp = F.normalize(inp, dim=-1)
if not advanced:
return rgbs[0], rgbs[1], rgbs[2], logging_str
if "AlignedThreeModelAttnNodes" == model_name:
return rgbs[0], rgbs[1], rgbs[2], logging_str
if advanced:
cluster_plots, norm_plots = [], []
for i in range(3):
eigvecs = all_eigvecs[i]
# add norm plot, cluster plot
start = time.time()
progress_start = 0.6
progress(progress_start, desc=f"Plotting Clusters Recursion #{i+1}")
h, w = features.shape[1], features.shape[2]
if torch.cuda.is_available():
images = images.cuda()
_images = reverse_transform_image(images, stablediffusion="stable" in model_name.lower())
cluster_images, eig_magnitude = make_cluster_plot_advanced(eigvecs, _images, h=h, w=w)
logging_str += f"Recursion #{i+1} plot time: {time.time() - start:.2f}s\n"
norm_images = []
vmin, vmax = eig_magnitude.min(), eig_magnitude.max()
eig_magnitude = (eig_magnitude - vmin) / (vmax - vmin)
eig_magnitude = eig_magnitude.cpu().numpy()
colormap = matplotlib.colormaps['Reds']
for i_image in range(eig_magnitude.shape[0]):
norm_image = colormap(eig_magnitude[i_image])
norm_images.append(torch.tensor(norm_image[..., :3]))
norm_images = to_pil_images(norm_images)
logging_str += f"Recursion #{i+1} Eigenvector Magnitude: [{vmin:.2f}, {vmax:.2f}]\n"
gr.Info(f"Recursion #{i+1} Eigenvector Magnitude:</br> Min: {vmin:.2f}, Max: {vmax:.2f}", duration=10)
cluster_plots.append(cluster_images)
norm_plots.append(norm_images)
return *rgbs, *norm_plots, *cluster_plots, logging_str
if old_school_ncut: # individual images
logging_str += "Running NCut for each image independently\n"
rgb = []
progress_start = 0.4
step_float = 0.6 / features.shape[0]
for i_image in range(features.shape[0]):
logging_str += f"Image #{i_image+1}\n"
feature = features[i_image]
_rgb, _logging_str, _ = compute_ncut(
feature[None],
num_eig=num_eig,
num_sample_ncut=30000,
affinity_focal_gamma=affinity_focal_gamma,
knn_ncut=1,
knn_tsne=10,
num_sample_tsne=300,
embedding_method=embedding_method,
embedding_metric=embedding_metric,
perplexity=perplexity,
n_neighbors=n_neighbors,
min_dist=min_dist,
sampling_method=sampling_method,
metric=ncut_metric,
indirect_connection=indirect_connection,
make_orthogonal=make_orthogonal,
progess_start=progress_start+step_float*i_image,
)
logging_str += _logging_str
rgb.append(_rgb[0])
return to_pil_images(rgb), logging_str
# ailgnedcut
if not directed:
rgb, _logging_str, eigvecs = compute_ncut(
features,
num_eig=num_eig,
num_sample_ncut=num_sample_ncut,
affinity_focal_gamma=affinity_focal_gamma,
knn_ncut=knn_ncut,
knn_tsne=knn_tsne,
num_sample_tsne=num_sample_tsne,
embedding_method=embedding_method,
embedding_metric=embedding_metric,
perplexity=perplexity,
n_neighbors=n_neighbors,
min_dist=min_dist,
sampling_method=sampling_method,
indirect_connection=indirect_connection,
make_orthogonal=make_orthogonal,
metric=ncut_metric,
)
if directed:
head_index_text = kwargs.get("head_index_text", None)
n_heads = features.shape[-2] # (batch, h, w, n_heads, d)
if head_index_text == 'all':
head_idx = torch.arange(n_heads)
else:
_idxs = head_index_text.split(",")
head_idx = torch.tensor([int(idx) for idx in _idxs])
features_A = features[:, :, :, head_idx, :]
features_B = features_B[:, :, :, head_idx, :]
rgb, _logging_str, eigvecs = compute_ncut_directed(
features_A,
features_B,
num_eig=num_eig,
num_sample_ncut=num_sample_ncut,
affinity_focal_gamma=affinity_focal_gamma,
knn_ncut=knn_ncut,
knn_tsne=knn_tsne,
num_sample_tsne=num_sample_tsne,
embedding_method=embedding_method,
embedding_metric=embedding_metric,
perplexity=perplexity,
n_neighbors=n_neighbors,
min_dist=min_dist,
sampling_method=sampling_method,
indirect_connection=False,
make_orthogonal=make_orthogonal,
metric=ncut_metric,
make_symmetric=kwargs.get("make_symmetric", None),
)
logging_str += _logging_str
if "AlignedThreeModelAttnNodes" == model_name:
# dirty patch for the alignedcut paper
start = time.time()
progress(0.6, desc="Plotting")
pil_images = []
for i_image in range(rgb.shape[0]):
_im = plot_one_image_36_grid(images[i_image], rgb[i_image])
pil_images.append(_im)
logging_str += f"plot time: {time.time() - start:.2f}s\n"
return pil_images, logging_str
if is_lisa == True:
# dirty patch for the LISA model
galleries = []
for i_prompt in range(len(lisa_prompts)):
_rgb = rgb[i_prompt]
galleries.append(to_pil_images(_rgb))
return *galleries, logging_str
rgb = dont_use_too_much_green(rgb)
if video_output:
progress(0.8, desc="Saving Video")
video_path = get_random_path()
video_cache.add_video(video_path)
pil_images_to_video(to_pil_images(rgb), video_path, fps=5)
return video_path, logging_str
cluster_images = None
if plot_clusters:
start = time.time()
progress_start = 0.6
progress(progress_start, desc="Plotting Clusters")
h, w = features.shape[1], features.shape[2]
if torch.cuda.is_available():
images = images.cuda()
_images = reverse_transform_image(images, stablediffusion="stable" in model_name.lower())
advanced = kwargs.get("advanced", False)
if advanced:
cluster_images, eig_magnitude = make_cluster_plot_advanced(eigvecs, _images, h=h, w=w)
else:
cluster_images, eig_magnitude = make_cluster_plot(eigvecs, _images, h=h, w=w, progess_start=progress_start, advanced=False)
logging_str += f"plot time: {time.time() - start:.2f}s\n"
norm_images = None
if alignedcut_eig_norm_plot:
norm_images = []
# eig_magnitude = torch.clamp(eig_magnitude, 0, 1)
vmin, vmax = eig_magnitude.min(), eig_magnitude.max()
eig_magnitude = (eig_magnitude - vmin) / (vmax - vmin)
eig_magnitude = eig_magnitude.cpu().numpy()
colormap = matplotlib.colormaps['Reds']
for i_image in range(eig_magnitude.shape[0]):
norm_image = colormap(eig_magnitude[i_image])
# norm_image = (norm_image[..., :3] * 255).astype(np.uint8)
# norm_images.append(Image.fromarray(norm_image))
norm_images.append(torch.tensor(norm_image[..., :3]))
norm_images = to_pil_images(norm_images)
logging_str += "Eigenvector Magnitude\n"
logging_str += f"Min: {vmin:.2f}, Max: {vmax:.2f}\n"
gr.Info(f"Eigenvector Magnitude:</br> Min: {vmin:.2f}, Max: {vmax:.2f}", duration=10)
return to_pil_images(rgb), cluster_images, norm_images, logging_str
def _ncut_run(*args, **kwargs):
n_ret = kwargs.pop("n_ret", 1)
# try:
# if torch.cuda.is_available():
# torch.cuda.empty_cache()
# ret = ncut_run(*args, **kwargs)
# if torch.cuda.is_available():
# torch.cuda.empty_cache()
# ret = list(ret)[:n_ret] + [ret[-1]]
# return ret
# except Exception as e:
# gr.Error(str(e))
# if torch.cuda.is_available():
# torch.cuda.empty_cache()
# return *(None for _ in range(n_ret)), "Error: " + str(e)
ret = ncut_run(*args, **kwargs)
ret = list(ret)[:n_ret] + [ret[-1]]
return ret
if USE_HUGGINGFACE_ZEROGPU:
@spaces.GPU(duration=30)
def quick_run(*args, **kwargs):
return _ncut_run(*args, **kwargs)
@spaces.GPU(duration=45)
def long_run(*args, **kwargs):
return _ncut_run(*args, **kwargs)
@spaces.GPU(duration=60)
def longer_run(*args, **kwargs):
return _ncut_run(*args, **kwargs)
@spaces.GPU(duration=120)
def super_duper_long_run(*args, **kwargs):
return _ncut_run(*args, **kwargs)
def cpu_run(*args, **kwargs):
return _ncut_run(*args, **kwargs)
if not USE_HUGGINGFACE_ZEROGPU:
def quick_run(*args, **kwargs):
return _ncut_run(*args, **kwargs)
def long_run(*args, **kwargs):
return _ncut_run(*args, **kwargs)
def longer_run(*args, **kwargs):
return _ncut_run(*args, **kwargs)
def super_duper_long_run(*args, **kwargs):
return _ncut_run(*args, **kwargs)
def cpu_run(*args, **kwargs):
return _ncut_run(*args, **kwargs)
def extract_video_frames(video_path, max_frames=100):
from decord import VideoReader
vr = VideoReader(video_path)
num_frames = len(vr)
if num_frames > max_frames:
gr.Warning(f"Video has {num_frames} frames. Only using {max_frames} frames. Evenly spaced.")
frame_idx = np.linspace(0, num_frames - 1, max_frames, dtype=int).tolist()
else:
frame_idx = list(range(num_frames))
frames = vr.get_batch(frame_idx).asnumpy()
# return as list of PIL images
return [(Image.fromarray(frames[i]), "") for i in range(frames.shape[0])]
def transform_image(image, resolution=(1024, 1024), stablediffusion=False):
image = image.convert('RGB').resize(resolution, Image.LANCZOS)
# Convert to torch tensor
image = torch.tensor(np.array(image).transpose(2, 0, 1)).float()
image = image / 255
# Normalize
if not stablediffusion:
mean = [0.485, 0.456, 0.406]
std = [0.229, 0.224, 0.225]
image = (image - torch.tensor(mean).view(3, 1, 1)) / torch.tensor(std).view(3, 1, 1)
if stablediffusion:
image = image * 2 - 1
return image
def reverse_transform_image(image, stablediffusion=False):
if stablediffusion:
image = (image + 1) / 2
else:
mean = torch.tensor([0.485, 0.456, 0.406]).view(3, 1, 1).to(image.device)
std = torch.tensor([0.229, 0.224, 0.225]).view(3, 1, 1).to(image.device)
image = image * std + mean
image = torch.clamp(image, 0, 1)
return image
def plot_one_image_36_grid(original_image, tsne_rgb_images):
mean = [0.485, 0.456, 0.406]
std = [0.229, 0.224, 0.225]
original_image = original_image * torch.tensor(std).view(3, 1, 1) + torch.tensor(mean).view(3, 1, 1)
original_image = torch.clamp(original_image, 0, 1)
fig = plt.figure(figsize=(20, 4))
grid = plt.GridSpec(3, 14, hspace=0.1, wspace=0.1)
ax1 = fig.add_subplot(grid[0:2, 0:2])
img = original_image.cpu().float().numpy().transpose(1, 2, 0)
def convert_and_pad_image(np_array, pad_size=20):
"""
Converts a NumPy array of shape (height, width, 3) to a PNG image
and pads the right and bottom sides with a transparent background.
Args:
np_array (numpy.ndarray): Input NumPy array of shape (height, width, 3)
pad_size (int, optional): Number of pixels to pad on the right and bottom sides. Default is 20.
Returns:
PIL.Image: Padded PNG image with transparent background
"""
# Convert NumPy array to PIL Image
img = Image.fromarray(np_array)
# Get the original size
width, height = img.size
# Create a new image with padding and transparent background
new_width = width + pad_size
new_height = height + pad_size
padded_img = Image.new('RGBA', (new_width, new_height), color=(255, 255, 255, 0))
# Paste the original image onto the padded image
padded_img.paste(img, (0, 0))
return padded_img
img = convert_and_pad_image((img*255).astype(np.uint8))
ax1.imshow(img)
ax1.axis('off')
model_names = ['CLIP', 'DINO', 'MAE']
for i_model, model_name in enumerate(model_names):
for i_layer in range(12):
ax = fig.add_subplot(grid[i_model, i_layer+2])
ax.imshow(tsne_rgb_images[i_layer+12*i_model].cpu().float().numpy())
ax.axis('off')
if i_model == 0:
ax.set_title(f'Layer{i_layer}', fontsize=16)
if i_layer == 0:
ax.text(-0.1, 0.5, model_name, va="center", ha="center", fontsize=16, transform=ax.transAxes, rotation=90,)
plt.tight_layout()
filename = uuid.uuid4()
filename = f"/tmp/{filename}.png"
plt.savefig(filename, bbox_inches='tight', pad_inches=0, dpi=100)
img = Image.open(filename)
img = img.convert("RGB")
img = copy.deepcopy(img)
os.remove(filename)
plt.close()
return img
def load_alignedthreemodel():
import sys
if "alignedthreeattn" not in sys.path:
for _ in range(3):
os.system("git clone https://huggingface.co/huzey/alignedthreeattn >> /dev/null 2>&1")
os.system("git -C alignedthreeattn pull >> /dev/null 2>&1")
# add to path
sys.path.append("alignedthreeattn")
from alignedthreeattn.alignedthreeattn_model import ThreeAttnNodes
align_weights = torch.load("alignedthreeattn/align_weights.pth")
model = ThreeAttnNodes(align_weights)
return model
try:
# pre-load the alignedthree model in case it fails to load
load_alignedthreemodel()
except Exception as e:
pass
promptable_diffusion_models = ["Diffusion(stabilityai/stable-diffusion-2)", "Diffusion(CompVis/stable-diffusion-v1-4)"]
promptable_segmentation_models = ["LISA(xinlai/LISA-7B-v1)"]
def run_fn(
images,
model_name="DiNO(dino_vitb8_448)",
layer=10,
num_eig=100,
node_type="block",
positive_prompt="",
negative_prompt="",
is_lisa=False,
lisa_prompt1="",
lisa_prompt2="",
lisa_prompt3="",
affinity_focal_gamma=0.5,
num_sample_ncut=10000,
knn_ncut=10,
ncut_indirect_connection=True,
ncut_make_orthogonal=False,
embedding_method="tsne_3d",
embedding_metric='euclidean',
num_sample_tsne=300,
knn_tsne=10,
perplexity=150,
n_neighbors=150,
min_dist=0.1,
sampling_method="QuickFPS",
ncut_metric="cosine",
old_school_ncut=False,
max_frames=100,
recursion=False,
recursion_l2_n_eigs=50,
recursion_l3_n_eigs=20,
recursion_metric="euclidean",
recursion_l1_gamma=0.5,
recursion_l2_gamma=0.5,
recursion_l3_gamma=0.5,
node_type2="k",
head_index_text='all',
make_symmetric=False,
n_ret=1,
plot_clusters=False,
alignedcut_eig_norm_plot=False,
advanced=False,
directed=False,
):
print(node_type2, head_index_text, make_symmetric)
progress=gr.Progress()
progress(0, desc="Starting")
if images is None:
gr.Warning("No images selected.")
return *(None for _ in range(n_ret)), "No images selected."
progress(0.05, desc="Processing Images")
video_output = False
if isinstance(images, str):
images = extract_video_frames(images, max_frames=max_frames)
video_output = True
if sampling_method == "QuickFPS":
sampling_method = "farthest"
# resize the images before acquiring GPU
if "AlignedThreeModelAttnNodes" == model_name:
# dirty patch for the alignedcut paper
resolution = (224, 224)
else:
resolution = RES_DICT[model_name]
images = [tup[0] for tup in images]
stablediffusion = True if "Diffusion" in model_name else False
images = [transform_image(image, resolution=resolution, stablediffusion=stablediffusion) for image in images]
images = torch.stack(images)
progress(0.1, desc="Downloading Model")
if is_lisa:
import subprocess
import sys
import importlib
gr.Warning("LISA model is not compatible with the current version of transformers. Please contact the LISA and Llava author for update.")
gr.Warning("This is a dirty patch for the LISA model. switch to the old version of transformers.")
gr.Warning("Not garanteed to work.")
# LISA and Llava is not compatible with the current version of transformers
# please contact the author for update
# this is a dirty patch for the LISA model
# pre-import the SD3 pipeline
from diffusers import StableDiffusion3Pipeline
# unloading the current transformers
for module in list(sys.modules.keys()):
if "transformers" in module:
del sys.modules[module]
def install_transformers_version(version, target_dir):
"""Install a specific version of transformers to a target directory."""
if not os.path.exists(target_dir):
os.makedirs(target_dir)
# Use subprocess to run the pip command
# subprocess.check_call([sys.executable, '-m', 'pip', 'install', f'transformers=={version}', '-t', target_dir])
os.system(f"{sys.executable} -m pip install transformers=={version} -t {target_dir} >> /dev/null 2>&1")
target_dir = '/tmp/lisa_transformers_v433'
if not os.path.exists(target_dir):
install_transformers_version('4.33.0', target_dir)
# Add the new version path to sys.path
sys.path.insert(0, target_dir)
transformers = importlib.import_module("transformers")
if not is_lisa:
import subprocess
import sys
import importlib
# remove the LISA model from the sys.path
if "/tmp/lisa_transformers_v433" in sys.path:
sys.path.remove("/tmp/lisa_transformers_v433")
transformers = importlib.import_module("transformers")
if "AlignedThreeModelAttnNodes" == model_name:
# dirty patch for the alignedcut paper
model = load_alignedthreemodel()
else:
model = load_model(model_name)
if "stable" in model_name.lower() and "diffusion" in model_name.lower():
model.timestep = layer
layer = 1
if model_name in promptable_diffusion_models:
model.positive_prompt = positive_prompt
model.negative_prompt = negative_prompt
kwargs = {
"model_name": model_name,
"layer": layer,
"num_eig": num_eig,
"node_type": node_type,
"affinity_focal_gamma": affinity_focal_gamma,
"num_sample_ncut": num_sample_ncut,
"knn_ncut": knn_ncut,
"embedding_method": embedding_method,
"embedding_metric": embedding_metric,
"num_sample_tsne": num_sample_tsne,
"knn_tsne": knn_tsne,
"perplexity": perplexity,
"n_neighbors": n_neighbors,
"min_dist": min_dist,
"sampling_method": sampling_method,
"ncut_metric": ncut_metric,
"indirect_connection": ncut_indirect_connection,
"make_orthogonal": ncut_make_orthogonal,
"old_school_ncut": old_school_ncut,
"recursion": recursion,
"recursion_l2_n_eigs": recursion_l2_n_eigs,
"recursion_l3_n_eigs": recursion_l3_n_eigs,
"recursion_metric": recursion_metric,
"recursion_l1_gamma": recursion_l1_gamma,
"recursion_l2_gamma": recursion_l2_gamma,
"recursion_l3_gamma": recursion_l3_gamma,
"video_output": video_output,
"lisa_prompt1": lisa_prompt1,
"lisa_prompt2": lisa_prompt2,
"lisa_prompt3": lisa_prompt3,
"is_lisa": is_lisa,
"n_ret": n_ret,
"plot_clusters": plot_clusters,
"alignedcut_eig_norm_plot": alignedcut_eig_norm_plot,
"advanced": advanced,
"directed": directed,
"node_type2": node_type2,
"head_index_text": head_index_text,
"make_symmetric": make_symmetric,
}
# print(kwargs)
try:
# try to aquiare GPU, can fail if the user is out of GPU quota
if old_school_ncut:
return super_duper_long_run(model, images, **kwargs)
if is_lisa:
return super_duper_long_run(model, images, **kwargs)
num_images = len(images)
if num_images >= 100:
return super_duper_long_run(model, images, **kwargs)
if 'diffusion' in model_name.lower():
return super_duper_long_run(model, images, **kwargs)
if recursion:
return longer_run(model, images, **kwargs)
if num_images >= 50:
return longer_run(model, images, **kwargs)
if old_school_ncut:
return longer_run(model, images, **kwargs)
if num_images >= 10:
return long_run(model, images, **kwargs)
if embedding_method == "UMAP":
if perplexity >= 250 or num_sample_tsne >= 500:
return longer_run(model, images, **kwargs)
return long_run(model, images, **kwargs)
if embedding_method == "t-SNE":
if perplexity >= 250 or num_sample_tsne >= 500:
return long_run(model, images, **kwargs)
return quick_run(model, images, **kwargs)
return quick_run(model, images, **kwargs)
except gr.Error as e:
# I assume this is a GPU quota error
info1 = 'Running out of HuggingFace GPU Quota?</br> Please try <a style="white-space: nowrap;text-underline-offset: 2px;color: var(--body-text-color)" href="https://ncut-pytorch.readthedocs.io/en/latest/demo/">Demo hosted at UPenn</a></br>'
info2 = 'Or try use the Python package that powers this app: <a style="white-space: nowrap;text-underline-offset: 2px;color: var(--body-text-color)" href="https://ncut-pytorch.readthedocs.io/en/latest/">ncut-pytorch</a>'
info = info1 + info2
message = "<b>HuggingFace: </b></br>" + e.message + "</br></br>---------</br>" + "<b>`ncut-pytorch` Developer: </b></br>" + info
raise gr.Error(message, duration=0)
import torch
from torch import nn
from torch.utils.data import Dataset, DataLoader
import pytorch_lightning as pl
# Custom Dataset
class TwoTensorDataset(Dataset):
def __init__(self, A, B):
self.A = A
self.B = B
def __len__(self):
return len(self.A)
def __getitem__(self, idx):
return self.A[idx], self.B[idx]
# MLP model
class MLP(pl.LightningModule):
def __init__(self, num_layer=3, width=512, lr=3e-4, fitting_steps=10000, seg_loss_lambda=1.0):
super().__init__()
layers = [nn.Linear(3, width), nn.GELU()]
for _ in range(num_layer - 1):
layers.append(nn.Linear(width, width))
layers.append(nn.GELU())
layers.append(nn.Linear(width, 3))
self.layers = nn.Sequential(*layers)
self.mse_loss = nn.MSELoss()
self.lr = lr
self.fitting_steps = fitting_steps
self.seg_loss_lambda = seg_loss_lambda
self.progress = gr.Progress()
def forward(self, x):
return self.layers(x)
def training_step(self, batch, batch_idx):
x, y = batch
y_hat = self.forward(x)
loss = self.mse_loss(y_hat, y)
# loss = torch.nn.functional.mse_loss(torch.log(y_hat), torch.log(y))
self.log("train_loss", loss)
# add segmentation constraint
bsz = x.shape[0]
sample_size = 1000
if bsz > sample_size:
idx = torch.randperm(bsz)[:sample_size]
x = x[idx]
y_hat = y_hat[idx]
old_dist = torch.pdist(x, p=2)
new_dist = torch.pdist(y_hat, p=2)
# seg_loss = torch.log((old_dist - new_dist)).pow(2).mean()
seg_loss = self.mse_loss(old_dist, new_dist)
self.log("seg_loss", seg_loss)
loss += seg_loss * self.seg_loss_lambda
step = self.global_step
if step % 100 == 0:
self.progress(step / self.fitting_steps, desc="Fitting MLP")
return loss
def predict_step(self, batch, batch_idx, dataloader_idx=None):
x = batch[0]
y_hat = self.forward(x)
return y_hat
def configure_optimizers(self):
optimizer = torch.optim.Adam(self.parameters(), lr=self.lr)
return optimizer
def fit_trans(rgb1, rgb2, num_layer=3, width=512, batch_size=256, lr=3e-4, fitting_steps=10000, fps_sample=4096, seg_loss_lambda=1.0):
A = rgb1.clone()
B = rgb2.clone()
# FPS sample on the data
from ncut_pytorch.ncut_pytorch import farthest_point_sampling
A_idx = farthest_point_sampling(A, fps_sample)
B_idx = farthest_point_sampling(B, fps_sample)
A_B_idx = np.concatenate([A_idx, B_idx])
A = A[A_B_idx]
B = B[A_B_idx]
from torch.utils.data import DataLoader, TensorDataset
# Dataset and DataLoader
dataset = TwoTensorDataset(A, B)
dataloader = DataLoader(dataset, batch_size=batch_size, shuffle=True)
# Initialize model and trainer
mlp = MLP(num_layer=num_layer, width=width, lr=lr, fitting_steps=fitting_steps, seg_loss_lambda=seg_loss_lambda)
trainer = pl.Trainer(
max_epochs=100000,
gpus=1,
max_steps=fitting_steps,
enable_checkpointing=False,
enable_progress_bar=False,
gradient_clip_val=1.0
)
# Create a DataLoader for tensor A
batch_size = 256 # Define your batch size
data_loader = DataLoader(TensorDataset(rgb1), batch_size=batch_size, shuffle=False)
# Train the model
trainer.fit(mlp, dataloader)
mlp.progress(0.99, desc="Applying MLP")
results = trainer.predict(mlp, data_loader)
A_transformed = torch.cat(results, dim=0)
return A_transformed
if USE_HUGGINGFACE_ZEROGPU:
@spaces.GPU(duration=60)
def _run_mlp_fit(*args, **kwargs):
return fit_trans(*args, **kwargs)
else:
def _run_mlp_fit(*args, **kwargs):
return fit_trans(*args, **kwargs)
def run_mlp_fit(input_gallery, target_gallery, num_layer=3, width=512, batch_size=256, lr=3e-4, fitting_steps=10000, fps_sample=4096, seg_loss_lambda=1.0):
# print("Fitting MLP")
# print("Target Gallery Length:", len(target_gallery))
# print("Input Gallery Length:", len(input_gallery))
if target_gallery is None or len(target_gallery) == 0:
raise gr.Error("No target images selected. Please use the Mark button to select the target images.")
if input_gallery is None or len(input_gallery) == 0:
raise gr.Error("No input images selected.")
def gallery_to_rgb(gallery):
images = [tup[0] for tup in gallery]
rgb = []
for image in images:
if isinstance(image, str):
image = Image.open(image)
image = image.convert('RGB')
image = torch.tensor(np.array(image)).float()
image = image / 255
rgb.append(image)
rgb = torch.stack(rgb)
shape = rgb.shape
rgb = rgb.reshape(-1, 3)
return rgb, shape
target_rgb, target_shape = gallery_to_rgb(target_gallery)
input_rgb, input_shape = gallery_to_rgb(input_gallery)
input_transformed = _run_mlp_fit(input_rgb, target_rgb, num_layer=num_layer, width=width, batch_size=batch_size, lr=lr,
fitting_steps=fitting_steps, fps_sample=fps_sample, seg_loss_lambda=seg_loss_lambda)
input_transformed = input_transformed.reshape(*input_shape)
pil_images = to_pil_images(input_transformed, resize=False)
return pil_images
def make_input_video_section():
# gr.Markdown('### Input Video')
input_gallery = gr.Video(value=None, label="Select video", elem_id="video-input", height="auto", show_share_button=False, interactive=True)
gr.Markdown('_image backbone model is used to extract features from each frame, NCUT is computed on all frames_')
max_frames_number = gr.Number(100, label="Max frames", elem_id="max_frames")
# max_frames_number = gr.Slider(1, 200, step=1, label="Max frames", value=100, elem_id="max_frames")
submit_button = gr.Button("🔴 RUN", elem_id="submit_button", variant='primary')
clear_images_button = gr.Button("🗑️Clear", elem_id='clear_button', variant='stop')
return input_gallery, submit_button, clear_images_button, max_frames_number
def load_dataset_images(is_advanced, dataset_name, num_images=10,
is_filter=False, filter_by_class_text="0,1,2",
is_random=False, seed=1):
progress = gr.Progress()
progress(0, desc="Loading Images")
if dataset_name == "EgoExo":
is_advanced = "Basic"
if is_advanced == "Basic":
gr.Info(f"Loaded images from EgoExo")
return default_images
try:
progress(0.5, desc="Downloading Dataset")
if 'EgoThink' in dataset_name:
dataset = load_dataset(dataset_name, 'Activity', trust_remote_code=True)
else:
dataset = load_dataset(dataset_name, trust_remote_code=True)
key = list(dataset.keys())[0]
dataset = dataset[key]
except Exception as e:
raise gr.Error(f"Error loading dataset {dataset_name}: {e}")
if num_images > len(dataset):
num_images = len(dataset)
if len(filter_by_class_text) == 0:
is_filter = False
if is_filter:
progress(0.8, desc="Filtering Images")
classes = [int(i) for i in filter_by_class_text.split(",")]
labels = np.array(dataset['label'])
unique_labels = np.unique(labels)
valid_classes = [i for i in classes if i in unique_labels]
invalid_classes = [i for i in classes if i not in unique_labels]
if len(invalid_classes) > 0:
gr.Warning(f"Classes {invalid_classes} not found in the dataset.")
if len(valid_classes) == 0:
raise gr.Error(f"Classes {classes} not found in the dataset.")
# shuffle each class
chunk_size = num_images // len(valid_classes)
image_idx = []
for i in valid_classes:
idx = np.where(labels == i)[0]
if is_random:
idx = np.random.RandomState(seed).choice(idx, chunk_size, replace=False)
else:
idx = idx[:chunk_size]
image_idx.extend(idx.tolist())
if not is_filter:
if is_random:
image_idx = np.random.RandomState(seed).choice(len(dataset), num_images, replace=False).tolist()
else:
image_idx = list(range(num_images))
key = 'image' if 'image' in dataset[0] else list(dataset[0].keys())[0]
images = [dataset[i][key] for i in image_idx]
gr.Info(f"Loaded {len(images)} images from {dataset_name}")
del dataset
if dataset_name in CENTER_CROP_DATASETS:
def center_crop_image(img):
# image: PIL image
w, h = img.size
min_hw = min(h, w)
# center crop
left = (w - min_hw) // 2
top = (h - min_hw) // 2
right = left + min_hw
bottom = top + min_hw
img = img.crop((left, top, right, bottom))
return img
images = [center_crop_image(image) for image in images]
return images
def load_and_append(existing_images, *args, **kwargs):
new_images = load_dataset_images(*args, **kwargs)
if new_images is None:
return existing_images
if len(new_images) == 0:
return existing_images
if existing_images is None:
existing_images = []
existing_images += new_images
gr.Info(f"Total images: {len(existing_images)}")
return existing_images
def make_input_images_section(rows=1, cols=3, height="auto", advanced=False, is_random=False, allow_download=False):
gr.Markdown('### Input Images')
input_gallery = gr.Gallery(value=None, label="Input images", show_label=True, elem_id="input_images", columns=[cols], rows=[rows], object_fit="contain", height=height, type="pil", show_share_button=False,
format="webp")
submit_button = gr.Button("🔴 RUN", elem_id="submit_button", variant='primary')
with gr.Row():
clear_images_button = gr.Button("🗑️ Clear", elem_id='clear_button', variant='stop')
clear_images_button.click(fn=lambda: gr.update(value=None), outputs=[input_gallery])
upload_button = gr.UploadButton(elem_id="upload_button", label="⬆️ Upload", variant='secondary', file_types=["image"], file_count="multiple")
def convert_to_pil_and_append(images, new_images):
if images is None:
images = []
if new_images is None:
return images
if isinstance(new_images, Image.Image):
images.append(new_images)
if isinstance(new_images, list):
images += [Image.open(new_image) for new_image in new_images]
if isinstance(new_images, str):
images.append(Image.open(new_images))
gr.Info(f"Total images: {len(images)}")
return images
upload_button.upload(convert_to_pil_and_append, inputs=[input_gallery, upload_button], outputs=[input_gallery])
if allow_download:
create_file_button, download_button = add_download_button(input_gallery, "input_images")
gr.Markdown('### Load Datasets')
advanced_radio = gr.Radio(["Basic", "Advanced"], label="Datasets Menu", value="Advanced" if advanced else "Basic", elem_id="advanced-radio", show_label=True)
with gr.Column() as basic_block:
# gr.Markdown('### Example Image Sets')
def make_example(name, images, dataset_name):
with gr.Row():
button = gr.Button("Load\n"+name, elem_id=f"example-{name}", elem_classes="small-button", variant='secondary', size="sm", scale=1, min_width=60)
gallery = gr.Gallery(value=images, label=name, show_label=True, columns=[3], rows=[1], interactive=False, height=80, scale=8, object_fit="cover", min_width=140, allow_preview=False)
button.click(fn=lambda: gr.update(value=load_dataset_images(True, dataset_name, 100, is_random=True, seed=42)), outputs=[input_gallery])
return gallery, button
example_items = [
("EgoExo", ['./images/egoexo1.jpg', './images/egoexo3.jpg', './images/egoexo2.jpg'], "EgoExo"),
("Ego", ['./images/egothink1.jpg', './images/egothink2.jpg', './images/egothink3.jpg'], "EgoThink/EgoThink"),
("Face", ['./images/face1.jpg', './images/face2.jpg', './images/face3.jpg'], "nielsr/CelebA-faces"),
("Pose", ['./images/pose1.jpg', './images/pose2.jpg', './images/pose3.jpg'], "sayakpaul/poses-controlnet-dataset"),
# ("CatDog", ['./images/catdog1.jpg', './images/catdog2.jpg', './images/catdog3.jpg'], "microsoft/cats_vs_dogs"),
# ("Bird", ['./images/bird1.jpg', './images/bird2.jpg', './images/bird3.jpg'], "Multimodal-Fatima/CUB_train"),
# ("ChestXray", ['./images/chestxray1.jpg', './images/chestxray2.jpg', './images/chestxray3.jpg'], "hongrui/mimic_chest_xray_v_1"),
("BrainMRI", ['./images/brain1.jpg', './images/brain2.jpg', './images/brain3.jpg'], "sartajbhuvaji/Brain-Tumor-Classification"),
("Kanji", ['./images/kanji1.jpg', './images/kanji2.jpg', './images/kanji3.jpg'], "yashvoladoddi37/kanjienglish"),
]
for name, images, dataset_name in example_items:
make_example(name, images, dataset_name)
with gr.Column() as advanced_block:
load_images_button = gr.Button("🔴 Load Images", elem_id="load-images-button", variant='primary')
# dataset_names = DATASET_NAMES
# dataset_classes = DATASET_CLASSES
dataset_categories = list(DATASETS.keys())
defualt_cat = dataset_categories[0]
def get_choices(cat):
return [tup[0] for tup in DATASETS[cat]]
defualt_choices = get_choices(defualt_cat)
with gr.Row():
dataset_radio = gr.Radio(dataset_categories, label="Dataset Category", value=defualt_cat, elem_id="dataset-radio", show_label=True, min_width=600)
# dataset_dropdown = gr.Dropdown(dataset_names, label="Dataset name", value="mrm8488/ImageNet1K-val", elem_id="dataset", min_width=300)
dataset_dropdown = gr.Dropdown(defualt_choices, label="Dataset name", value=defualt_choices[0], elem_id="dataset", min_width=400)
dataset_radio.change(fn=lambda x: gr.update(choices=get_choices(x), value=get_choices(x)[0]), inputs=dataset_radio, outputs=dataset_dropdown)
# num_images_slider = gr.Number(10, label="Number of images", elem_id="num_images")
num_images_slider = gr.Slider(1, 1000, step=1, label="Number of images", value=10, elem_id="num_images", min_width=200)
if not is_random:
filter_by_class_checkbox = gr.Checkbox(label="Filter by class", value=True, elem_id="filter_by_class_checkbox")
filter_by_class_text = gr.Textbox(label="Class to select", value="0,33,99", elem_id="filter_by_class_text", info=f"e.g. `0,1,2`. (1000 classes)", visible=True)
# is_random_checkbox = gr.Checkbox(label="Random shuffle", value=False, elem_id="random_seed_checkbox")
# random_seed_slider = gr.Slider(0, 1000, step=1, label="Random seed", value=1, elem_id="random_seed", visible=False)
is_random_checkbox = gr.Checkbox(label="Random shuffle", value=True, elem_id="random_seed_checkbox")
random_seed_slider = gr.Slider(0, 1000, step=1, label="Random seed", value=1, elem_id="random_seed", visible=True)
if is_random:
filter_by_class_checkbox = gr.Checkbox(label="Filter by class", value=False, elem_id="filter_by_class_checkbox")
filter_by_class_text = gr.Textbox(label="Class to select", value="0,33,99", elem_id="filter_by_class_text", info=f"e.g. `0,1,2`. (1000 classes)", visible=False)
is_random_checkbox = gr.Checkbox(label="Random shuffle", value=True, elem_id="random_seed_checkbox")
random_seed_slider = gr.Slider(0, 1000, step=1, label="Random seed", value=42, elem_id="random_seed", visible=True)
# add functionality, save and load images to profile
with gr.Accordion("Saved Image Profiles", open=False) as profile_accordion:
with gr.Row():
profile_text = gr.Textbox(label="Profile name", placeholder="Type here: Profile name to save/load/delete", elem_id="profile-name", scale=6, show_label=False)
list_profiles_button = gr.Button("📋 List", elem_id="list-profile-button", variant='secondary', scale=3)
with gr.Row():
save_profile_button = gr.Button("💾 Save", elem_id="save-profile-button", variant='secondary')
load_profile_button = gr.Button("📂 Load", elem_id="load-profile-button", variant='secondary')
delete_profile_button = gr.Button("🗑️ Delete", elem_id="delete-profile-button", variant='secondary')
class OnDiskProfiles:
def __init__(self, profile_dir="demo_profiles"):
if not os.path.exists(profile_dir):
os.makedirs(profile_dir)
self.profile_dir = profile_dir
def list_profiles(self):
profiles = os.listdir(self.profile_dir)
# remove hidden files
profiles = [p for p in profiles if not p.startswith(".")]
if len(profiles) == 0:
return "No profiles found."
profile_text = "</br>".join(profiles)
n_files = len(profiles)
profile_text = f"Number of profiles: {n_files}</br>---------</br>" + profile_text
return profile_text
def save_profile(self, profile_name, images):
profile_path = os.path.join(self.profile_dir, profile_name)
if os.path.exists(profile_path):
raise gr.Error(f"Profile {profile_name} already exists.")
with open(profile_path, "wb") as f:
pickle.dump(images, f)
gr.Info(f"Profile {profile_name} saved.")
return profile_path
def load_profile(self, profile_name, existing_images):
profile_path = os.path.join(self.profile_dir, profile_name)
if not os.path.exists(profile_path):
raise gr.Error(f"Profile {profile_name} not found.")
with open(profile_path, "rb") as f:
images = pickle.load(f)
gr.Info(f"Profile {profile_name} loaded.")
if existing_images is None:
existing_images = []
return existing_images + images
def delete_profile(self, profile_name):
profile_path = os.path.join(self.profile_dir, profile_name)
os.remove(profile_path)
gr.Info(f"Profile {profile_name} deleted.")
return profile_path
home_dir = os.path.expanduser("~")
defualt_dir = os.path.join(home_dir, ".cache")
cache_dir = os.environ.get("DEMO_PROFILE_CACHE_DIR", defualt_dir)
cache_dir = os.path.join(cache_dir, "demo_profiles")
on_disk_profiles = OnDiskProfiles(cache_dir)
save_profile_button.click(fn=lambda name, images: on_disk_profiles.save_profile(name, images), inputs=[profile_text, input_gallery])
load_profile_button.click(fn=lambda name, existing_images: gr.update(value=on_disk_profiles.load_profile(name, existing_images)), inputs=[profile_text, input_gallery], outputs=[input_gallery])
delete_profile_button.click(fn=lambda name: on_disk_profiles.delete_profile(name), inputs=profile_text)
list_profiles_button.click(fn=lambda: gr.Info(on_disk_profiles.list_profiles(), duration=0))
if advanced:
advanced_block.visible = True
basic_block.visible = False
else:
advanced_block.visible = False
basic_block.visible = True
# change visibility
advanced_radio.change(fn=lambda x: gr.update(visible=x=="Advanced"), inputs=advanced_radio, outputs=[advanced_block])
advanced_radio.change(fn=lambda x: gr.update(visible=x=="Basic"), inputs=advanced_radio, outputs=[basic_block])
def find_num_classes(dataset_name):
num_classes = None
for cat, datasets in DATASETS.items():
datasets = [tup[0] for tup in datasets]
if dataset_name in datasets:
num_classes = DATASETS[cat][datasets.index(dataset_name)][1]
break
return num_classes
def change_filter_options(dataset_name):
num_classes = find_num_classes(dataset_name)
if num_classes is None:
return (gr.Checkbox(label="Filter by class", value=False, elem_id="filter_by_class_checkbox", visible=False),
gr.Textbox(label="Class to select", value="0,1,2", elem_id="filter_by_class_text", info="e.g. `0,1,2`. This dataset has no class label", visible=False))
return (gr.Checkbox(label="Filter by class", value=True, elem_id="filter_by_class_checkbox", visible=True),
gr.Textbox(label="Class to select", value="0,1,2", elem_id="filter_by_class_text", info=f"e.g. `0,1,2`. ({num_classes} classes)", visible=True))
dataset_dropdown.change(fn=change_filter_options, inputs=dataset_dropdown, outputs=[filter_by_class_checkbox, filter_by_class_text])
def change_filter_by_class(is_filter, dataset_name):
num_classes = find_num_classes(dataset_name)
return gr.Textbox(label="Class to select", value="0,1,2", elem_id="filter_by_class_text", info=f"e.g. `0,1,2`. ({num_classes} classes)", visible=is_filter)
filter_by_class_checkbox.change(fn=change_filter_by_class, inputs=[filter_by_class_checkbox, dataset_dropdown], outputs=filter_by_class_text)
def change_random_seed(is_random):
return gr.Slider(0, 1000, step=1, label="Random seed", value=1, elem_id="random_seed", visible=is_random)
is_random_checkbox.change(fn=change_random_seed, inputs=is_random_checkbox, outputs=random_seed_slider)
load_images_button.click(load_and_append,
inputs=[input_gallery, advanced_radio, dataset_dropdown, num_images_slider,
filter_by_class_checkbox, filter_by_class_text,
is_random_checkbox, random_seed_slider],
outputs=[input_gallery])
return input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button
# def random_rotate_rgb_gallery(images):
# if images is None or len(images) == 0:
# gr.Warning("No images selected.")
# return []
# # read webp images
# images = [Image.open(image[0]).convert("RGB") for image in images]
# images = [np.array(image).astype(np.float32) for image in images]
# images = np.stack(images)
# images = torch.tensor(images) / 255
# position = np.random.choice([1, 2, 4, 5, 6])
# images = rotate_rgb_cube(images, position)
# images = to_pil_images(images, resize=False)
# return images
def protect_original_image_in_plot(original_image, rotated_images):
plot_h, plot_w = 332, 1542
image_h, image_w = original_image.shape[1], original_image.shape[2]
if not (plot_h == image_h and plot_w == image_w):
return rotated_images
protection_w = 190
rotated_images[:, :, :protection_w] = original_image[:, :, :protection_w]
return rotated_images
def sequence_rotate_rgb_gallery(images):
if images is None or len(images) == 0:
gr.Warning("No images selected.")
return []
# read webp images
images = [Image.open(image[0]).convert("RGB") for image in images]
images = [np.array(image).astype(np.float32) for image in images]
images = np.stack(images)
images = torch.tensor(images) / 255
original_images = images.clone()
rotation_matrix = torch.tensor([[0, 1, 0], [0, 0, 1], [1, 0, 0]]).float()
images = images @ rotation_matrix
images = protect_original_image_in_plot(original_images, images)
images = to_pil_images(images, resize=False)
return images
def flip_rgb_gallery(images, axis=0):
if images is None or len(images) == 0:
gr.Warning("No images selected.")
return []
# read webp images
images = [Image.open(image[0]).convert("RGB") for image in images]
images = [np.array(image).astype(np.float32) for image in images]
images = np.stack(images)
images = torch.tensor(images) / 255
original_images = images.clone()
images = 1 - images
images = protect_original_image_in_plot(original_images, images)
images = to_pil_images(images, resize=False)
return images
def add_rotate_flip_buttons(output_gallery):
with gr.Row():
rotate_button = gr.Button("🔄 Rotate", elem_id="rotate_button", variant='secondary')
rotate_button.click(sequence_rotate_rgb_gallery, inputs=[output_gallery], outputs=[output_gallery])
flip_button = gr.Button("🔃 Flip", elem_id="flip_button", variant='secondary')
flip_button.click(flip_rgb_gallery, inputs=[output_gallery], outputs=[output_gallery])
return rotate_button, flip_button
def add_download_button(gallery, filename_prefix="output"):
def make_3x5_plot(images):
plot_list = []
# Split the list of images into chunks of 15
chunks = [images[i:i + 15] for i in range(0, len(images), 15)]
for chunk in chunks:
fig, axs = plt.subplots(3, 4, figsize=(12, 9))
for ax in axs.flatten():
ax.axis("off")
for ax, img in zip(axs.flatten(), chunk):
img = img.convert("RGB")
ax.imshow(img)
plt.tight_layout(h_pad=0.5, w_pad=0.3)
# Generate a unique filename
filename = uuid.uuid4()
tmp_path = f"/tmp/{filename}.png"
# Save the plot to the temporary file
plt.savefig(tmp_path, bbox_inches='tight', dpi=144)
# Open the saved image
img = Image.open(tmp_path)
img = img.convert("RGB")
img = copy.deepcopy(img)
# Remove the temporary file
os.remove(tmp_path)
plot_list.append(img)
plt.close()
return plot_list
def delete_file_after_delay(file_path, delay):
def delete_file():
if os.path.exists(file_path):
os.remove(file_path)
timer = threading.Timer(delay, delete_file)
timer.start()
def create_zip_file(images, filename_prefix=filename_prefix):
if images is None or len(images) == 0:
gr.Warning("No images selected.")
return None
gr.Info("Creating zip file for download...")
images = [image[0] for image in images]
if isinstance(images[0], str):
images = [Image.open(image) for image in images]
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
zip_filename = f"/tmp/gallery_download/{filename_prefix}_{timestamp}.zip"
os.makedirs(os.path.dirname(zip_filename), exist_ok=True)
plots = make_3x5_plot(images)
with zipfile.ZipFile(zip_filename, 'w') as zipf:
# Create a temporary directory to store images and plots
temp_dir = f"/tmp/gallery_download/images/{uuid.uuid4()}"
os.makedirs(temp_dir)
try:
# Save images to the temporary directory
for i, img in enumerate(images):
img = img.convert("RGB")
img_path = os.path.join(temp_dir, f"single_{i:04d}.jpg")
img.save(img_path)
zipf.write(img_path, f"single_{i:04d}.jpg")
# Save plots to the temporary directory
for i, plot in enumerate(plots):
plot = plot.convert("RGB")
plot_path = os.path.join(temp_dir, f"grid_{i:04d}.jpg")
plot.save(plot_path)
zipf.write(plot_path, f"grid_{i:04d}.jpg")
finally:
# Clean up the temporary directory
for file in os.listdir(temp_dir):
os.remove(os.path.join(temp_dir, file))
os.rmdir(temp_dir)
# Schedule the deletion of the zip file after 24 hours (86400 seconds)
delete_file_after_delay(zip_filename, 86400)
gr.Info(f"File is ready for download: {os.path.basename(zip_filename)}")
return gr.update(value=zip_filename, interactive=True)
with gr.Row():
create_file_button = gr.Button("📦 Pack", elem_id="create_file_button", variant='secondary')
download_button = gr.DownloadButton(label="📥 Download", value=None, variant='secondary', elem_id="download_button", interactive=False)
create_file_button.click(create_zip_file, inputs=[gallery], outputs=[download_button])
def warn_on_click(filename):
if filename is None:
gr.Warning("No file to download, please `📦 Pack` first.")
interactive = filename is not None
return gr.update(interactive=interactive)
download_button.click(warn_on_click, inputs=[download_button], outputs=[download_button])
return create_file_button, download_button
def make_output_images_section():
gr.Markdown('### Output Images')
output_gallery = gr.Gallery(format='png', value=[], label="NCUT Embedding", show_label=True, elem_id="ncut", columns=[3], rows=[1], object_fit="contain", height="auto", show_share_button=True, interactive=False)
add_rotate_flip_buttons(output_gallery)
return output_gallery
def make_parameters_section(is_lisa=False, model_ratio=True):
gr.Markdown("### Parameters <a style='color: #0044CC;' href='https://ncut-pytorch.readthedocs.io/en/latest/how_to_get_better_segmentation/' target='_blank'>Help</a>")
from ncut_pytorch.backbone import list_models, get_demo_model_names
model_names = list_models()
model_names = sorted(model_names)
def get_filtered_model_names(name):
return [m for m in model_names if name.lower() in m.lower()]
def get_default_model_name(name):
lst = get_filtered_model_names(name)
if len(lst) > 1:
return lst[1]
return lst[0]
if is_lisa:
model_dropdown = gr.Dropdown(["LISA(xinlai/LISA-7B-v1)"], label="Backbone", value="LISA(xinlai/LISA-7B-v1)", elem_id="model_name")
layer_slider = gr.Slider(1, 6, step=1, label="LISA decoder: Layer index", value=6, elem_id="layer", visible=False)
layer_names = ["dec_0_input", "dec_0_attn", "dec_0_block", "dec_1_input", "dec_1_attn", "dec_1_block"]
positive_prompt = gr.Textbox(label="Prompt (Positive)", elem_id="prompt", placeholder="e.g. 'a photo of Gibson Les Pual guitar'", visible=False)
negative_prompt = gr.Textbox(label="Prompt (Negative)", elem_id="prompt", placeholder="e.g. 'a photo from egocentric view'", visible=False)
node_type_dropdown = gr.Dropdown(layer_names, label="LISA (SAM) decoder: Layer and Node", value="dec_1_block", elem_id="node_type")
else:
model_radio = gr.Radio(["CLIP", "DiNO", "Diffusion", "ImageNet", "MAE", "SAM"], label="Backbone", value="DiNO", elem_id="model_radio", show_label=True, visible=model_ratio)
model_dropdown = gr.Dropdown(get_filtered_model_names("DiNO"), label="", value="DiNO(dino_vitb8_448)", elem_id="model_name", show_label=False)
model_radio.change(fn=lambda x: gr.update(choices=get_filtered_model_names(x), value=get_default_model_name(x)), inputs=model_radio, outputs=[model_dropdown])
layer_slider = gr.Slider(1, 12, step=1, label="Backbone: Layer index", value=10, elem_id="layer")
positive_prompt = gr.Textbox(label="Prompt (Positive)", elem_id="prompt", placeholder="e.g. 'a photo of Gibson Les Pual guitar'")
positive_prompt.visible = False
negative_prompt = gr.Textbox(label="Prompt (Negative)", elem_id="prompt", placeholder="e.g. 'a photo from egocentric view'")
negative_prompt.visible = False
node_type_dropdown = gr.Dropdown(["attn: attention output", "mlp: mlp output", "block: sum of residual"], label="Backbone: Layer type", value="block: sum of residual", elem_id="node_type", info="which feature to take from each layer?")
num_eig_slider = gr.Slider(1, 1000, step=1, label="NCUT: Number of eigenvectors", value=100, elem_id="num_eig", info='increase for smaller clusters')
def change_layer_slider(model_name):
# SD2, UNET
if "stable" in model_name.lower() and "diffusion" in model_name.lower():
from ncut_pytorch.backbone import SD_KEY_DICT
default_layer = 'up_2_resnets_1_block' if 'diffusion-3' not in model_name else 'block_23'
return (gr.Slider(1, 49, step=1, label="Diffusion: Timestep (Noise)", value=5, elem_id="layer", visible=True, info="Noise level, 50 is max noise"),
gr.Dropdown(SD_KEY_DICT[model_name], label="Diffusion: Layer and Node", value=default_layer, elem_id="node_type", info="U-Net (v1, v2) or DiT (v3)"))
if model_name == "LISSL(xinlai/LISSL-7B-v1)":
layer_names = ["dec_0_input", "dec_0_attn", "dec_0_block", "dec_1_input", "dec_1_attn", "dec_1_block"]
default_layer = "dec_1_block"
return (gr.Slider(1, 6, step=1, label="LISA decoder: Layer index", value=6, elem_id="layer", visible=False, info=""),
gr.Dropdown(layer_names, label="LISA decoder: Layer and Node", value=default_layer, elem_id="node_type"))
layer_dict = LAYER_DICT
if model_name in layer_dict:
value = layer_dict[model_name]
return (gr.Slider(1, value, step=1, label="Backbone: Layer index", value=value, elem_id="layer", visible=True, info=""),
gr.Dropdown(["attn: attention output", "mlp: mlp output", "block: sum of residual"], label="Backbone: Layer type", value="block: sum of residual", elem_id="node_type", info="which feature to take from each layer?"))
else:
value = 12
return (gr.Slider(1, value, step=1, label="Backbone: Layer index", value=value, elem_id="layer", visible=True, info=""),
gr.Dropdown(["attn: attention output", "mlp: mlp output", "block: sum of residual"], label="Backbone: Layer type", value="block: sum of residual", elem_id="node_type", info="which feature to take from each layer?"))
model_dropdown.change(fn=change_layer_slider, inputs=model_dropdown, outputs=[layer_slider, node_type_dropdown])
def change_prompt_text(model_name):
if model_name in promptable_diffusion_models:
return (gr.Textbox(label="Prompt (Positive)", elem_id="prompt", placeholder="e.g. 'a photo of Gibson Les Pual guitar'", visible=True),
gr.Textbox(label="Prompt (Negative)", elem_id="prompt", placeholder="e.g. 'a photo from egocentric view'", visible=True))
return (gr.Textbox(label="Prompt (Positive)", elem_id="prompt", placeholder="e.g. 'a photo of Gibson Les Pual guitar'", visible=False),
gr.Textbox(label="Prompt (Negative)", elem_id="prompt", placeholder="e.g. 'a photo from egocentric view'", visible=False))
model_dropdown.change(fn=change_prompt_text, inputs=model_dropdown, outputs=[positive_prompt, negative_prompt])
with gr.Accordion("Advanced Parameters: NCUT", open=False):
gr.Markdown("<a href='https://ncut-pytorch.readthedocs.io/en/latest/how_to_get_better_segmentation/' target='_blank'>Docs: How to Get Better Segmentation</a>")
affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="NCUT: Affinity focal gamma", value=0.5, elem_id="affinity_focal_gamma", info="decrease for shaper segmentation")
num_sample_ncut_slider = gr.Slider(100, 50000, step=100, label="NCUT: num_sample", value=10000, elem_id="num_sample_ncut", info="Nyström approximation")
# sampling_method_dropdown = gr.Dropdown(["QuickFPS", "random"], label="NCUT: Sampling method", value="QuickFPS", elem_id="sampling_method", info="Nyström approximation")
sampling_method_dropdown = gr.Radio(["QuickFPS", "random"], label="NCUT: Sampling method", value="QuickFPS", elem_id="sampling_method")
# ncut_metric_dropdown = gr.Dropdown(["euclidean", "cosine"], label="NCUT: Distance metric", value="cosine", elem_id="ncut_metric")
ncut_metric_dropdown = gr.Radio(["euclidean", "cosine"], label="NCUT: Distance metric", value="cosine", elem_id="ncut_metric")
ncut_knn_slider = gr.Slider(1, 100, step=1, label="NCUT: KNN", value=10, elem_id="knn_ncut", info="Nyström approximation")
ncut_indirect_connection = gr.Checkbox(label="indirect_connection", value=True, elem_id="ncut_indirect_connection", info="Add indirect connection to the sub-sampled graph")
ncut_make_orthogonal = gr.Checkbox(label="make_orthogonal", value=False, elem_id="ncut_make_orthogonal", info="Apply post-hoc eigenvectors orthogonalization")
with gr.Accordion("Advanced Parameters: Visualization", open=False):
# embedding_method_dropdown = gr.Dropdown(["tsne_3d", "umap_3d", "umap_sphere", "tsne_2d", "umap_2d"], label="Coloring method", value="tsne_3d", elem_id="embedding_method")
embedding_method_dropdown = gr.Radio(["tsne_3d", "umap_3d", "umap_sphere", "tsne_2d", "umap_2d"], label="Coloring method", value="tsne_3d", elem_id="embedding_method")
# embedding_metric_dropdown = gr.Dropdown(["euclidean", "cosine"], label="t-SNE/UMAP metric", value="euclidean", elem_id="embedding_metric")
embedding_metric_dropdown = gr.Radio(["euclidean", "cosine"], label="t-SNE/UMAP: metric", value="euclidean", elem_id="embedding_metric")
num_sample_tsne_slider = gr.Slider(100, 10000, step=100, label="t-SNE/UMAP: num_sample", value=300, elem_id="num_sample_tsne", info="Nyström approximation")
knn_tsne_slider = gr.Slider(1, 100, step=1, label="t-SNE/UMAP: KNN", value=10, elem_id="knn_tsne", info="Nyström approximation")
perplexity_slider = gr.Slider(10, 1000, step=10, label="t-SNE: perplexity", value=150, elem_id="perplexity")
n_neighbors_slider = gr.Slider(10, 1000, step=10, label="UMAP: n_neighbors", value=150, elem_id="n_neighbors")
min_dist_slider = gr.Slider(0.1, 1, step=0.1, label="UMAP: min_dist", value=0.1, elem_id="min_dist")
return [model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt]
custom_css = """
#unlock_button {
all: unset !important;
}
.form:has(#unlock_button) {
all: unset !important;
}
"""
demo = gr.Blocks(
theme=gr.themes.Base(spacing_size='md', text_size='lg', primary_hue='blue', neutral_hue='slate', secondary_hue='pink'),
# fill_width=False,
# title="ncut-pytorch",
css=custom_css,
)
with demo:
with gr.Tab('AlignedCut'):
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section()
num_images_slider.value = 30
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information", autofocus=False, autoscroll=False)
with gr.Column(scale=5, min_width=200):
output_gallery = make_output_images_section()
cluster_gallery = gr.Gallery(value=[], label="Clusters", show_label=True, elem_id="clusters", columns=[2], rows=[2], object_fit="contain", height="auto", show_share_button=True, preview=False, interactive=False)
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section()
false_placeholder = gr.Checkbox(label="False", value=False, elem_id="false_placeholder", visible=False)
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
submit_button.click(
partial(run_fn, n_ret=2, plot_clusters=True),
inputs=[
input_gallery, model_dropdown, layer_slider, num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown
],
outputs=[output_gallery, cluster_gallery, logging_text],
api_name="API_AlignedCut",
scroll_to_output=True,
)
with gr.Tab('AlignedCut (Advanced)', visible=False) as tab_alignedcut_advanced:
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section(allow_download=True)
num_images_slider.value = 100
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information", autofocus=False, autoscroll=False, lines=20)
with gr.Column(scale=5, min_width=200):
output_gallery = make_output_images_section()
add_download_button(output_gallery, "ncut_embed")
norm_gallery = gr.Gallery(value=[], label="Eigenvector Magnitude", show_label=True, elem_id="eig_norm", columns=[3], rows=[1], object_fit="contain", height="auto", show_share_button=True, preview=False, interactive=False)
add_download_button(norm_gallery, "eig_norm")
cluster_gallery = gr.Gallery(value=[], label="Clusters", show_label=True, elem_id="clusters", columns=[2], rows=[4], object_fit="contain", height="auto", show_share_button=True, preview=False, interactive=False)
add_download_button(cluster_gallery, "clusters")
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section()
num_eig_slider.value = 100
false_placeholder = gr.Checkbox(label="False", value=False, elem_id="false_placeholder", visible=False)
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
submit_button.click(
partial(run_fn, n_ret=3, plot_clusters=True, alignedcut_eig_norm_plot=True, advanced=True),
inputs=[
input_gallery, model_dropdown, layer_slider, num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown
],
outputs=[output_gallery, cluster_gallery, norm_gallery, logging_text],
scroll_to_output=True,
)
with gr.Tab('NCut'):
gr.Markdown('#### NCut (Legacy), not aligned, no Nyström approximation')
gr.Markdown('Each image is solved independently, <em>color is <b>not</b> aligned across images</em>')
gr.Markdown('---')
gr.Markdown('<p style="text-align: center;"><b>NCut vs. AlignedCut</b></p>')
with gr.Row():
with gr.Column(scale=5, min_width=200):
gr.Markdown('#### Pros')
gr.Markdown('- Easy Solution. Use less eigenvectors.')
gr.Markdown('- Exact solution. No Nyström approximation.')
with gr.Column(scale=5, min_width=200):
gr.Markdown('#### Cons')
gr.Markdown('- Not aligned. Distance is not preserved across images. No pseudo-labeling or correspondence.')
gr.Markdown('- Poor complexity scaling. Unable to handle large number of pixels.')
gr.Markdown('---')
with gr.Row():
with gr.Column(scale=5, min_width=200):
gr.Markdown(' ')
with gr.Column(scale=5, min_width=200):
gr.Markdown('<em>color is <b>not</b> aligned across images</em> 👇')
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section()
with gr.Column(scale=5, min_width=200):
output_gallery = make_output_images_section()
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section()
old_school_ncut_checkbox = gr.Checkbox(label="Old school NCut", value=True, elem_id="old_school_ncut")
invisible_list = [old_school_ncut_checkbox, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
num_sample_tsne_slider, knn_tsne_slider, sampling_method_dropdown, ncut_metric_dropdown]
for item in invisible_list:
item.visible = False
# logging text box
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information")
false_placeholder = gr.Checkbox(label="False", value=False, elem_id="false_placeholder", visible=False)
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
submit_button.click(
run_fn,
inputs=[
input_gallery, model_dropdown, layer_slider, num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown,
old_school_ncut_checkbox
],
outputs=[output_gallery, logging_text],
api_name="API_NCut",
)
with gr.Tab('Recursive Cut'):
gr.Markdown('NCUT can be applied recursively, the eigenvectors from previous iteration is the input for the next iteration NCUT. ')
gr.Markdown('__Recursive NCUT__ can amplify or weaken the connections, depending on the `affinity_focal_gamma` setting, please see [Documentation](https://ncut-pytorch.readthedocs.io/en/latest/how_to_get_better_segmentation/#recursive-ncut)')
gr.Markdown('---')
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section()
num_images_slider.value = 100
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information")
with gr.Column(scale=5, min_width=200):
gr.Markdown('### Output (Recursion #1)')
l1_gallery = gr.Gallery(format='png', value=[], label="Recursion #1", show_label=True, elem_id="ncut_l1", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
add_rotate_flip_buttons(l1_gallery)
with gr.Column(scale=5, min_width=200):
gr.Markdown('### Output (Recursion #2)')
l2_gallery = gr.Gallery(format='png', value=[], label="Recursion #2", show_label=True, elem_id="ncut_l2", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
add_rotate_flip_buttons(l2_gallery)
with gr.Column(scale=5, min_width=200):
gr.Markdown('### Output (Recursion #3)')
l3_gallery = gr.Gallery(format='png', value=[], label="Recursion #3", show_label=True, elem_id="ncut_l3", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
add_rotate_flip_buttons(l3_gallery)
with gr.Row():
with gr.Column(scale=5, min_width=200):
with gr.Accordion("➡️ Recursion config", open=True):
l1_num_eig_slider = gr.Slider(1, 1000, step=1, label="Recursion #1: N eigenvectors", value=100, elem_id="l1_num_eig")
l2_num_eig_slider = gr.Slider(1, 1000, step=1, label="Recursion #2: N eigenvectors", value=50, elem_id="l2_num_eig")
l3_num_eig_slider = gr.Slider(1, 1000, step=1, label="Recursion #3: N eigenvectors", value=50, elem_id="l3_num_eig")
metric_dropdown = gr.Dropdown(["euclidean", "cosine"], label="Recursion distance metric", value="cosine", elem_id="recursion_metric")
l1_affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="Recursion #1: Affinity focal gamma", value=0.5, elem_id="recursion_l1_gamma")
l2_affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="Recursion #2: Affinity focal gamma", value=0.5, elem_id="recursion_l2_gamma")
l3_affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="Recursion #3: Affinity focal gamma", value=0.5, elem_id="recursion_l3_gamma")
with gr.Column(scale=5, min_width=200):
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section()
num_eig_slider.visible = False
affinity_focal_gamma_slider.visible = False
true_placeholder = gr.Checkbox(label="True placeholder", value=True, elem_id="true_placeholder")
true_placeholder.visible = False
false_placeholder = gr.Checkbox(label="False placeholder", value=False, elem_id="false_placeholder")
false_placeholder.visible = False
number_placeholder = gr.Number(0, label="Number placeholder", elem_id="number_placeholder")
number_placeholder.visible = False
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
submit_button.click(
partial(run_fn, n_ret=3),
inputs=[
input_gallery, model_dropdown, layer_slider, l1_num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown,
false_placeholder, number_placeholder, true_placeholder,
l2_num_eig_slider, l3_num_eig_slider, metric_dropdown,
l1_affinity_focal_gamma_slider, l2_affinity_focal_gamma_slider, l3_affinity_focal_gamma_slider
],
outputs=[l1_gallery, l2_gallery, l3_gallery, logging_text],
api_name="API_RecursiveCut"
)
with gr.Tab('Recursive Cut (Advanced)', visible=False) as tab_recursivecut_advanced:
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section(allow_download=True)
num_images_slider.value = 100
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information", lines=20)
with gr.Column(scale=5, min_width=200):
gr.Markdown('### Output (Recursion #1)')
l1_gallery = gr.Gallery(format='png', value=[], label="Recursion #1", show_label=True, elem_id="ncut_l1", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
add_rotate_flip_buttons(l1_gallery)
add_download_button(l1_gallery, "ncut_embed_recur1")
l1_norm_gallery = gr.Gallery(value=[], label="Recursion #1 Eigenvector Magnitude", show_label=True, elem_id="eig_norm", columns=[3], rows=[1], object_fit="contain", height="auto", show_share_button=True, preview=False, interactive=False)
add_download_button(l1_norm_gallery, "eig_norm_recur1")
l1_cluster_gallery = gr.Gallery(value=[], label="Recursion #1 Clusters", show_label=True, elem_id="clusters", columns=[2], rows=[4], object_fit="contain", height='auto', show_share_button=True, preview=False, interactive=False)
add_download_button(l1_cluster_gallery, "clusters_recur1")
with gr.Column(scale=5, min_width=200):
gr.Markdown('### Output (Recursion #2)')
l2_gallery = gr.Gallery(format='png', value=[], label="Recursion #2", show_label=True, elem_id="ncut_l2", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
add_rotate_flip_buttons(l2_gallery)
add_download_button(l2_gallery, "ncut_embed_recur2")
l2_norm_gallery = gr.Gallery(value=[], label="Recursion #2 Eigenvector Magnitude", show_label=True, elem_id="eig_norm", columns=[3], rows=[1], object_fit="contain", height="auto", show_share_button=True, preview=False, interactive=False)
add_download_button(l2_norm_gallery, "eig_norm_recur2")
l2_cluster_gallery = gr.Gallery(value=[], label="Recursion #2 Clusters", show_label=True, elem_id="clusters", columns=[2], rows=[4], object_fit="contain", height='auto', show_share_button=True, preview=False, interactive=False)
add_download_button(l2_cluster_gallery, "clusters_recur2")
with gr.Column(scale=5, min_width=200):
gr.Markdown('### Output (Recursion #3)')
l3_gallery = gr.Gallery(format='png', value=[], label="Recursion #3", show_label=True, elem_id="ncut_l3", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
add_rotate_flip_buttons(l3_gallery)
add_download_button(l3_gallery, "ncut_embed_recur3")
l3_norm_gallery = gr.Gallery(value=[], label="Recursion #3 Eigenvector Magnitude", show_label=True, elem_id="eig_norm", columns=[3], rows=[1], object_fit="contain", height="auto", show_share_button=True, preview=False, interactive=False)
add_download_button(l3_norm_gallery, "eig_norm_recur3")
l3_cluster_gallery = gr.Gallery(value=[], label="Recursion #3 Clusters", show_label=True, elem_id="clusters", columns=[2], rows=[4], object_fit="contain", height='auto', show_share_button=True, preview=False, interactive=False)
add_download_button(l3_cluster_gallery, "clusters_recur3")
with gr.Row():
with gr.Column(scale=5, min_width=200):
with gr.Accordion("➡️ Recursion config", open=True):
l1_num_eig_slider = gr.Slider(1, 1000, step=1, label="Recursion #1: N eigenvectors", value=100, elem_id="l1_num_eig")
l2_num_eig_slider = gr.Slider(1, 1000, step=1, label="Recursion #2: N eigenvectors", value=50, elem_id="l2_num_eig")
l3_num_eig_slider = gr.Slider(1, 1000, step=1, label="Recursion #3: N eigenvectors", value=50, elem_id="l3_num_eig")
metric_dropdown = gr.Dropdown(["euclidean", "cosine"], label="Recursion distance metric", value="cosine", elem_id="recursion_metric")
l1_affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="Recursion #1: Affinity focal gamma", value=0.5, elem_id="recursion_l1_gamma")
l2_affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="Recursion #2: Affinity focal gamma", value=0.5, elem_id="recursion_l2_gamma")
l3_affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="Recursion #3: Affinity focal gamma", value=0.5, elem_id="recursion_l3_gamma")
with gr.Column(scale=5, min_width=200):
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section()
num_eig_slider.visible = False
affinity_focal_gamma_slider.visible = False
true_placeholder = gr.Checkbox(label="True placeholder", value=True, elem_id="true_placeholder")
true_placeholder.visible = False
false_placeholder = gr.Checkbox(label="False placeholder", value=False, elem_id="false_placeholder")
false_placeholder.visible = False
number_placeholder = gr.Number(0, label="Number placeholder", elem_id="number_placeholder")
number_placeholder.visible = False
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
submit_button.click(
partial(run_fn, n_ret=9, advanced=True),
inputs=[
input_gallery, model_dropdown, layer_slider, l1_num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown,
false_placeholder, number_placeholder, true_placeholder,
l2_num_eig_slider, l3_num_eig_slider, metric_dropdown,
l1_affinity_focal_gamma_slider, l2_affinity_focal_gamma_slider, l3_affinity_focal_gamma_slider
],
outputs=[l1_gallery, l2_gallery, l3_gallery, l1_norm_gallery, l2_norm_gallery, l3_norm_gallery, l1_cluster_gallery, l2_cluster_gallery, l3_cluster_gallery, logging_text],
)
with gr.Tab('Video'):
with gr.Row():
with gr.Column(scale=5, min_width=200):
video_input_gallery, submit_button, clear_video_button, max_frame_number = make_input_video_section()
with gr.Column(scale=5, min_width=200):
video_output_gallery = gr.Video(value=None, label="NCUT Embedding", elem_id="ncut", height="auto", show_share_button=False)
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section()
num_sample_tsne_slider.value = 1000
perplexity_slider.value = 500
n_neighbors_slider.value = 500
knn_tsne_slider.value = 20
# logging text box
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information")
clear_video_button.click(lambda x: (None, None), outputs=[video_input_gallery, video_output_gallery])
place_holder_false = gr.Checkbox(label="Place holder", value=False, elem_id="place_holder_false")
place_holder_false.visible = False
false_placeholder = gr.Checkbox(label="False", value=False, elem_id="false_placeholder", visible=False)
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
submit_button.click(
run_fn,
inputs=[
video_input_gallery, model_dropdown, layer_slider, num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown,
place_holder_false, max_frame_number
],
outputs=[video_output_gallery, logging_text],
api_name="API_VideoCut",
)
with gr.Tab('Text'):
try:
from app_text import make_demo
except ImportError:
print("Debugging")
from draft_gradio_app_text import make_demo
make_demo()
with gr.Tab('Vision-Language'):
gr.Markdown('[LISA](https://arxiv.org/pdf/2308.00692) is a vision-language model. Input a text prompt and image, LISA generate segmentation masks.')
gr.Markdown('In the mask decoder layers, LISA updates the image features w.r.t. the text prompt')
gr.Markdown('This page aims to see how the text prompt affects the image features')
gr.Markdown('---')
gr.Markdown('<p style="text-align: center;">Color is <b>aligned</b> across 3 prompts. NCUT is computed on the concatenated features from 3 prompts.</p>')
with gr.Row():
with gr.Column(scale=5, min_width=200):
gr.Markdown('### Output (Prompt #1)')
l1_gallery = gr.Gallery(format='png', value=[], label="Prompt #1", show_label=False, elem_id="ncut_p1", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
prompt1 = gr.Textbox(label="Input Prompt #1", elem_id="prompt1", value="where is the person, include the clothes, don't include the guitar and chair", lines=3)
with gr.Column(scale=5, min_width=200):
gr.Markdown('### Output (Prompt #2)')
l2_gallery = gr.Gallery(format='png', value=[], label="Prompt #2", show_label=False, elem_id="ncut_p2", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
prompt2 = gr.Textbox(label="Input Prompt #2", elem_id="prompt2", value="where is the Gibson Les Pual guitar", lines=3)
with gr.Column(scale=5, min_width=200):
gr.Markdown('### Output (Prompt #3)')
l3_gallery = gr.Gallery(format='png', value=[], label="Prompt #3", show_label=False, elem_id="ncut_p3", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
prompt3 = gr.Textbox(label="Input Prompt #3", elem_id="prompt3", value="where is the floor", lines=3)
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section()
with gr.Column(scale=5, min_width=200):
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section(is_lisa=True)
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information")
galleries = [l1_gallery, l2_gallery, l3_gallery]
true_placeholder = gr.Checkbox(label="True placeholder", value=True, elem_id="true_placeholder", visible=False)
submit_button.click(
partial(run_fn, n_ret=len(galleries)),
inputs=[
input_gallery, model_dropdown, layer_slider, num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
true_placeholder, prompt1, prompt2, prompt3,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown
],
outputs=galleries + [logging_text],
)
with gr.Tab('Model Aligned'):
gr.Markdown('This page reproduce the results from the paper [AlignedCut](https://arxiv.org/abs/2406.18344)')
gr.Markdown('---')
gr.Markdown('**Features are aligned across models and layers.** A linear alignment transform is trained for each model/layer, learning signal comes from 1) fMRI brain activation and 2) segmentation preserving eigen-constraints.')
gr.Markdown('NCUT is computed on the concatenated graph of all models, layers, and images. Color is **aligned** across all models and layers.')
gr.Markdown('')
gr.Markdown("To see a good pattern, you will need to load 100~1000 images. 100 images need 10sec for RTX4090. Running out of HuggingFace GPU Quota? Try [Demo](https://ncut-pytorch.readthedocs.io/en/latest/demo/) hosted at UPenn")
gr.Markdown('---')
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section()
num_images_slider.value = 100
with gr.Column(scale=5, min_width=200):
output_gallery = make_output_images_section()
gr.Markdown('### TIP1: use the `full-screen` button, and use `arrow keys` to navigate')
gr.Markdown('---')
gr.Markdown('Model: CLIP(ViT-B-16/openai), DiNOv2reg(dinov2_vitb14_reg), MAE(vit_base)')
gr.Markdown('Layer type: attention output (attn), without sum of residual')
gr.Markdown('### TIP2: for large image set, please increase the `num_sample` for t-SNE and NCUT')
gr.Markdown('---')
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section(model_ratio=False)
model_dropdown.value = "AlignedThreeModelAttnNodes"
model_dropdown.visible = False
layer_slider.visible = False
node_type_dropdown.visible = False
num_sample_ncut_slider.value = 10000
num_sample_tsne_slider.value = 1000
# logging text box
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information")
false_placeholder = gr.Checkbox(label="False", value=False, elem_id="false_placeholder", visible=False)
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
submit_button.click(
run_fn,
inputs=[
input_gallery, model_dropdown, layer_slider, num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown
],
# outputs=galleries + [logging_text],
outputs=[output_gallery, logging_text],
)
with gr.Tab('Model Aligned (Advanced)', visible=False) as tab_model_aligned_advanced:
gr.Markdown('This page reproduce the results from the paper [AlignedCut](https://arxiv.org/abs/2406.18344)')
gr.Markdown('---')
gr.Markdown('**Features are aligned across models and layers.** A linear alignment transform is trained for each model/layer, learning signal comes from 1) fMRI brain activation and 2) segmentation preserving eigen-constraints.')
gr.Markdown('NCUT is computed on the concatenated graph of all models, layers, and images. Color is **aligned** across all models and layers.')
gr.Markdown('')
gr.Markdown("To see a good pattern, you will need to load 100~1000 images. 100 images need 10sec for RTX4090. Running out of HuggingFace GPU Quota? Try [Demo](https://ncut-pytorch.readthedocs.io/en/latest/demo/) hosted at UPenn")
gr.Markdown('---')
# with gr.Row():
# with gr.Column(scale=5, min_width=200):
# gr.Markdown('### Output (Recursion #1)')
# l1_gallery = gr.Gallery(format='png', value=[], label="Recursion #1", show_label=False, elem_id="ncut_l1", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
# add_output_images_buttons(l1_gallery)
# with gr.Column(scale=5, min_width=200):
# gr.Markdown('### Output (Recursion #2)')
# l2_gallery = gr.Gallery(format='png', value=[], label="Recursion #2", show_label=False, elem_id="ncut_l2", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
# add_output_images_buttons(l2_gallery)
# with gr.Column(scale=5, min_width=200):
# gr.Markdown('### Output (Recursion #3)')
# l3_gallery = gr.Gallery(format='png', value=[], label="Recursion #3", show_label=False, elem_id="ncut_l3", columns=[3], rows=[5], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
# add_output_images_buttons(l3_gallery)
gr.Markdown('### Output (Recursion #1)')
l1_gallery = gr.Gallery(format='png', value=[], label="Recursion #1", show_label=True, elem_id="ncut_l1", columns=[100], rows=[1], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False, preview=True)
add_rotate_flip_buttons(l1_gallery)
add_download_button(l1_gallery, "modelaligned_recur1")
gr.Markdown('### Output (Recursion #2)')
l2_gallery = gr.Gallery(format='png', value=[], label="Recursion #2", show_label=True, elem_id="ncut_l2", columns=[100], rows=[1], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False, preview=True)
add_rotate_flip_buttons(l2_gallery)
add_download_button(l2_gallery, "modelaligned_recur2")
gr.Markdown('### Output (Recursion #3)')
l3_gallery = gr.Gallery(format='png', value=[], label="Recursion #3", show_label=True, elem_id="ncut_l3", columns=[100], rows=[1], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False, preview=True)
add_rotate_flip_buttons(l3_gallery)
add_download_button(l3_gallery, "modelaligned_recur3")
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section(allow_download=True)
num_images_slider.value = 100
with gr.Column(scale=5, min_width=200):
with gr.Accordion("➡️ Recursion config", open=True):
l1_num_eig_slider = gr.Slider(1, 1000, step=1, label="Recursion #1: N eigenvectors", value=100, elem_id="l1_num_eig")
l2_num_eig_slider = gr.Slider(1, 1000, step=1, label="Recursion #2: N eigenvectors", value=50, elem_id="l2_num_eig")
l3_num_eig_slider = gr.Slider(1, 1000, step=1, label="Recursion #3: N eigenvectors", value=50, elem_id="l3_num_eig")
metric_dropdown = gr.Dropdown(["euclidean", "cosine"], label="Recursion distance metric", value="cosine", elem_id="recursion_metric")
l1_affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="Recursion #1: Affinity focal gamma", value=0.5, elem_id="recursion_l1_gamma")
l2_affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="Recursion #2: Affinity focal gamma", value=0.5, elem_id="recursion_l2_gamma")
l3_affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="Recursion #3: Affinity focal gamma", value=0.5, elem_id="recursion_l3_gamma")
gr.Markdown('---')
gr.Markdown('Model: CLIP(ViT-B-16/openai), DiNOv2reg(dinov2_vitb14_reg), MAE(vit_base)')
gr.Markdown('Layer type: attention output (attn), without sum of residual')
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section(model_ratio=False)
num_eig_slider.visible = False
affinity_focal_gamma_slider.visible = False
model_dropdown.value = "AlignedThreeModelAttnNodes"
model_dropdown.visible = False
layer_slider.visible = False
node_type_dropdown.visible = False
num_sample_ncut_slider.value = 10000
num_sample_tsne_slider.value = 1000
# logging text box
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information")
true_placeholder = gr.Checkbox(label="True placeholder", value=True, elem_id="true_placeholder")
true_placeholder.visible = False
false_placeholder = gr.Checkbox(label="False placeholder", value=False, elem_id="false_placeholder")
false_placeholder.visible = False
number_placeholder = gr.Number(0, label="Number placeholder", elem_id="number_placeholder")
number_placeholder.visible = False
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
submit_button.click(
partial(run_fn, n_ret=3, advanced=True),
inputs=[
input_gallery, model_dropdown, layer_slider, l1_num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown,
false_placeholder, number_placeholder, true_placeholder,
l2_num_eig_slider, l3_num_eig_slider, metric_dropdown,
l1_affinity_focal_gamma_slider, l2_affinity_focal_gamma_slider, l3_affinity_focal_gamma_slider
],
outputs=[l1_gallery, l2_gallery, l3_gallery, logging_text],
)
with gr.Tab('Compare Models'):
def add_one_model(i_model=1):
with gr.Column(scale=5, min_width=200) as col:
gr.Markdown(f'### Output Images')
output_gallery = gr.Gallery(format='png', value=[], label="NCUT Embedding", show_label=False, elem_id=f"ncut{i_model}", columns=[3], rows=[1], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
submit_button = gr.Button("🔴 RUN", elem_id=f"submit_button{i_model}", variant='primary')
add_rotate_flip_buttons(output_gallery)
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section()
# logging text box
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information")
false_placeholder = gr.Checkbox(label="False", value=False, elem_id="false_placeholder", visible=False)
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
submit_button.click(
run_fn,
inputs=[
input_gallery, model_dropdown, layer_slider, num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown
],
outputs=[output_gallery, logging_text]
)
return col
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section()
submit_button.visible = False
for i in range(3):
add_one_model()
# Create rows and buttons in a loop
rows = []
buttons = []
for i in range(4):
row = gr.Row(visible=False)
rows.append(row)
with row:
for j in range(4):
with gr.Column(scale=5, min_width=200):
add_one_model()
button = gr.Button("➕ Add Compare", elem_id=f"add_button_{i}", visible=False if i > 0 else True, scale=3)
buttons.append(button)
if i > 0:
# Reveal the current row and next button
buttons[i - 1].click(fn=lambda x: gr.update(visible=True), outputs=row)
buttons[i - 1].click(fn=lambda x: gr.update(visible=True), outputs=button)
# Hide the current button
buttons[i - 1].click(fn=lambda x: gr.update(visible=False), outputs=buttons[i - 1])
# Last button only reveals the last row and hides itself
buttons[-1].click(fn=lambda x: gr.update(visible=True), outputs=rows[-1])
buttons[-1].click(fn=lambda x: gr.update(visible=False), outputs=buttons[-1])
with gr.Tab('Compare Models (Advanced)', visible=False) as tab_compare_models_advanced:
target_images = gr.State([])
input_images = gr.State([])
def add_mlp_fitting_buttons(output_gallery, mlp_gallery, target_images=target_images, input_images=input_images):
with gr.Row():
# mark_as_target_button = gr.Button("mark target", elem_id=f"mark_as_target_button_{output_gallery.elem_id}", variant='secondary')
# mark_as_input_button = gr.Button("mark input", elem_id=f"mark_as_input_button_{output_gallery.elem_id}", variant='secondary')
mark_as_target_button = gr.Button("🎯 Mark Target", elem_id=f"mark_as_target_button_{output_gallery.elem_id}", variant='secondary')
fit_to_target_button = gr.Button("🔴 [MLP] Fit", elem_id=f"fit_to_target_button_{output_gallery.elem_id}", variant='primary')
def mark_fn(images, text="target"):
if images is None:
raise gr.Error("No images selected")
if len(images) == 0:
raise gr.Error("No images selected")
num_images = len(images)
gr.Info(f"Marked {num_images} images as {text}")
images = [(Image.open(tup[0]), []) for tup in images]
return images
mark_as_target_button.click(partial(mark_fn, text="target"), inputs=[output_gallery], outputs=[target_images])
# mark_as_input_button.click(partial(mark_fn, text="input"), inputs=[output_gallery], outputs=[input_images])
with gr.Accordion("➡️ MLP Parameters", open=False):
num_layers_slider = gr.Slider(2, 10, step=1, label="Number of Layers", value=3, elem_id=f"num_layers_slider_{output_gallery.elem_id}")
width_slider = gr.Slider(128, 4096, step=128, label="Width", value=512, elem_id=f"width_slider_{output_gallery.elem_id}")
batch_size_slider = gr.Slider(32, 4096, step=32, label="Batch Size", value=128, elem_id=f"batch_size_slider_{output_gallery.elem_id}")
lr_slider = gr.Slider(1e-6, 1, step=1e-6, label="Learning Rate", value=3e-4, elem_id=f"lr_slider_{output_gallery.elem_id}")
fitting_steps_slider = gr.Slider(1000, 100000, step=1000, label="Fitting Steps", value=30000, elem_id=f"fitting_steps_slider_{output_gallery.elem_id}")
fps_sample_slider = gr.Slider(128, 50000, step=128, label="FPS Sample", value=10240, elem_id=f"fps_sample_slider_{output_gallery.elem_id}")
segmentation_loss_lambda_slider = gr.Slider(0, 100, step=0.01, label="Segmentation Preserving Loss Lambda", value=1, elem_id=f"segmentation_loss_lambda_slider_{output_gallery.elem_id}")
fit_to_target_button.click(
run_mlp_fit,
inputs=[output_gallery, target_images, num_layers_slider, width_slider, batch_size_slider, lr_slider, fitting_steps_slider, fps_sample_slider, segmentation_loss_lambda_slider],
outputs=[mlp_gallery],
)
def add_one_model(i_model=1):
with gr.Column(scale=5, min_width=200) as col:
gr.Markdown(f'### Output Images')
output_gallery = gr.Gallery(format='png', value=[], label="NCUT Embedding", show_label=True, elem_id=f"ncut{i_model}", columns=[3], rows=[1], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
submit_button = gr.Button("🔴 RUN", elem_id=f"submit_button{i_model}", variant='primary')
add_rotate_flip_buttons(output_gallery)
add_download_button(output_gallery, f"ncut_embed")
mlp_gallery = gr.Gallery(format='png', value=[], label="MLP color align", show_label=True, elem_id=f"mlp{i_model}", columns=[3], rows=[1], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
add_mlp_fitting_buttons(output_gallery, mlp_gallery)
add_download_button(mlp_gallery, f"mlp_color_align")
norm_gallery = gr.Gallery(value=[], label="Eigenvector Magnitude", show_label=True, elem_id=f"eig_norm{i_model}", columns=[3], rows=[1], object_fit="contain", height="auto", show_share_button=True, preview=False, interactive=False)
add_download_button(norm_gallery, f"eig_norm")
cluster_gallery = gr.Gallery(value=[], label="Clusters", show_label=True, elem_id=f"clusters{i_model}", columns=[2], rows=[4], object_fit="contain", height="auto", show_share_button=True, preview=False, interactive=False)
add_download_button(cluster_gallery, f"clusters")
[
model_dropdown, layer_slider, node_type_dropdown, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section()
# logging text box
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information")
false_placeholder = gr.Checkbox(label="False", value=False, elem_id="false_placeholder", visible=False)
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
submit_button.click(
partial(run_fn, n_ret=3, plot_clusters=True, alignedcut_eig_norm_plot=True, advanced=True),
inputs=[
input_gallery, model_dropdown, layer_slider, num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown
],
outputs=[output_gallery, cluster_gallery, norm_gallery, logging_text]
)
output_gallery.change(lambda x: gr.update(value=x), inputs=[output_gallery], outputs=[mlp_gallery])
return output_gallery
galleries = []
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section(allow_download=True)
submit_button.visible = False
for i in range(3):
g = add_one_model()
galleries.append(g)
# Create rows and buttons in a loop
rows = []
buttons = []
for i in range(4):
row = gr.Row(visible=False)
rows.append(row)
with row:
for j in range(4):
with gr.Column(scale=5, min_width=200):
g = add_one_model()
galleries.append(g)
button = gr.Button("➕ Add Compare", elem_id=f"add_button_{i}", visible=False if i > 0 else True, scale=3)
buttons.append(button)
if i > 0:
# Reveal the current row and next button
buttons[i - 1].click(fn=lambda x: gr.update(visible=True), outputs=row)
buttons[i - 1].click(fn=lambda x: gr.update(visible=True), outputs=button)
# Hide the current button
buttons[i - 1].click(fn=lambda x: gr.update(visible=False), outputs=buttons[i - 1])
# Last button only reveals the last row and hides itself
buttons[-1].click(fn=lambda x: gr.update(visible=True), outputs=rows[-1])
buttons[-1].click(fn=lambda x: gr.update(visible=False), outputs=buttons[-1])
with gr.Tab('Directed (experimental)', visible=True) as tab_directed_ncut:
target_images = gr.State([])
input_images = gr.State([])
def add_mlp_fitting_buttons(output_gallery, mlp_gallery, target_images=target_images, input_images=input_images):
with gr.Row():
# mark_as_target_button = gr.Button("mark target", elem_id=f"mark_as_target_button_{output_gallery.elem_id}", variant='secondary')
# mark_as_input_button = gr.Button("mark input", elem_id=f"mark_as_input_button_{output_gallery.elem_id}", variant='secondary')
mark_as_target_button = gr.Button("🎯 Mark Target", elem_id=f"mark_as_target_button_{output_gallery.elem_id}", variant='secondary')
fit_to_target_button = gr.Button("🔴 [MLP] Fit", elem_id=f"fit_to_target_button_{output_gallery.elem_id}", variant='primary')
def mark_fn(images, text="target"):
if images is None:
raise gr.Error("No images selected")
if len(images) == 0:
raise gr.Error("No images selected")
num_images = len(images)
gr.Info(f"Marked {num_images} images as {text}")
images = [(Image.open(tup[0]), []) for tup in images]
return images
mark_as_target_button.click(partial(mark_fn, text="target"), inputs=[output_gallery], outputs=[target_images])
# mark_as_input_button.click(partial(mark_fn, text="input"), inputs=[output_gallery], outputs=[input_images])
with gr.Accordion("➡️ MLP Parameters", open=False):
num_layers_slider = gr.Slider(2, 10, step=1, label="Number of Layers", value=3, elem_id=f"num_layers_slider_{output_gallery.elem_id}")
width_slider = gr.Slider(128, 4096, step=128, label="Width", value=512, elem_id=f"width_slider_{output_gallery.elem_id}")
batch_size_slider = gr.Slider(32, 4096, step=32, label="Batch Size", value=128, elem_id=f"batch_size_slider_{output_gallery.elem_id}")
lr_slider = gr.Slider(1e-6, 1, step=1e-6, label="Learning Rate", value=3e-4, elem_id=f"lr_slider_{output_gallery.elem_id}")
fitting_steps_slider = gr.Slider(1000, 100000, step=1000, label="Fitting Steps", value=30000, elem_id=f"fitting_steps_slider_{output_gallery.elem_id}")
fps_sample_slider = gr.Slider(128, 50000, step=128, label="FPS Sample", value=10240, elem_id=f"fps_sample_slider_{output_gallery.elem_id}")
segmentation_loss_lambda_slider = gr.Slider(0, 100, step=0.01, label="Segmentation Preserving Loss Lambda", value=1, elem_id=f"segmentation_loss_lambda_slider_{output_gallery.elem_id}")
fit_to_target_button.click(
run_mlp_fit,
inputs=[output_gallery, target_images, num_layers_slider, width_slider, batch_size_slider, lr_slider, fitting_steps_slider, fps_sample_slider, segmentation_loss_lambda_slider],
outputs=[mlp_gallery],
)
def make_parameters_section_2model(model_ratio=True):
gr.Markdown("### Parameters <a style='color: #0044CC;' href='https://ncut-pytorch.readthedocs.io/en/latest/how_to_get_better_segmentation/' target='_blank'>Help</a>")
from ncut_pytorch.backbone import list_models, get_demo_model_names
model_names = list_models()
model_names = sorted(model_names)
# only CLIP DINO MAE is implemented for q k v
ok_models = ["CLIP(ViT", "DiNO(", "MAE("]
model_names = [m for m in model_names if any(ok in m for ok in ok_models)]
def get_filtered_model_names(name):
return [m for m in model_names if name.lower() in m.lower()]
def get_default_model_name(name):
lst = get_filtered_model_names(name)
if len(lst) > 1:
return lst[1]
return lst[0]
model_radio = gr.Radio(["CLIP", "DiNO", "MAE"], label="Backbone", value="DiNO", elem_id="model_radio", show_label=True, visible=model_ratio)
model_dropdown = gr.Dropdown(get_filtered_model_names("DiNO"), label="", value="DiNO(dino_vitb8_448)", elem_id="model_name", show_label=False)
model_radio.change(fn=lambda x: gr.update(choices=get_filtered_model_names(x), value=get_default_model_name(x)), inputs=model_radio, outputs=[model_dropdown])
layer_slider = gr.Slider(1, 12, step=1, label="Backbone: Layer index", value=10, elem_id="layer")
positive_prompt = gr.Textbox(label="Prompt (Positive)", elem_id="prompt", placeholder="e.g. 'a photo of Gibson Les Pual guitar'")
positive_prompt.visible = False
negative_prompt = gr.Textbox(label="Prompt (Negative)", elem_id="prompt", placeholder="e.g. 'a photo from egocentric view'")
negative_prompt.visible = False
node_type_dropdown = gr.Dropdown(['q', 'k', 'v'],
label="Left-side Node Type", value="q", elem_id="node_type", info="In directed case, left-side SVD eigenvector is taken")
node_type_dropdown2 = gr.Dropdown(['q', 'k', 'v'],
label="Right-side Node Type", value="k", elem_id="node_type2")
head_index_text = gr.Textbox(value='all', label="Head Index", elem_id="head_index", type="text", info="which attention heads to use, comma separated, e.g. 0,1,2")
make_symmetric = gr.Checkbox(label="Make Symmetric", value=False, elem_id="make_symmetric", info="make the graph symmetric by A = (A + A.T) / 2")
num_eig_slider = gr.Slider(1, 1000, step=1, label="NCUT: Number of eigenvectors", value=100, elem_id="num_eig", info='increase for smaller clusters')
def change_layer_slider(model_name):
# SD2, UNET
if "stable" in model_name.lower() and "diffusion" in model_name.lower():
from ncut_pytorch.backbone import SD_KEY_DICT
default_layer = 'up_2_resnets_1_block' if 'diffusion-3' not in model_name else 'block_23'
return (gr.Slider(1, 49, step=1, label="Diffusion: Timestep (Noise)", value=5, elem_id="layer", visible=True, info="Noise level, 50 is max noise"),
gr.Dropdown(SD_KEY_DICT[model_name], label="Diffusion: Layer and Node", value=default_layer, elem_id="node_type", info="U-Net (v1, v2) or DiT (v3)"))
if model_name == "LISSL(xinlai/LISSL-7B-v1)":
layer_names = ["dec_0_input", "dec_0_attn", "dec_0_block", "dec_1_input", "dec_1_attn", "dec_1_block"]
default_layer = "dec_1_block"
return (gr.Slider(1, 6, step=1, label="LISA decoder: Layer index", value=6, elem_id="layer", visible=False, info=""),
gr.Dropdown(layer_names, label="LISA decoder: Layer and Node", value=default_layer, elem_id="node_type"))
layer_dict = LAYER_DICT
if model_name in layer_dict:
value = layer_dict[model_name]
return gr.Slider(1, value, step=1, label="Backbone: Layer index", value=value, elem_id="layer", visible=True, info="")
else:
value = 12
return gr.Slider(1, value, step=1, label="Backbone: Layer index", value=value, elem_id="layer", visible=True, info="")
model_dropdown.change(fn=change_layer_slider, inputs=model_dropdown, outputs=layer_slider)
def change_prompt_text(model_name):
if model_name in promptable_diffusion_models:
return (gr.Textbox(label="Prompt (Positive)", elem_id="prompt", placeholder="e.g. 'a photo of Gibson Les Pual guitar'", visible=True),
gr.Textbox(label="Prompt (Negative)", elem_id="prompt", placeholder="e.g. 'a photo from egocentric view'", visible=True))
return (gr.Textbox(label="Prompt (Positive)", elem_id="prompt", placeholder="e.g. 'a photo of Gibson Les Pual guitar'", visible=False),
gr.Textbox(label="Prompt (Negative)", elem_id="prompt", placeholder="e.g. 'a photo from egocentric view'", visible=False))
model_dropdown.change(fn=change_prompt_text, inputs=model_dropdown, outputs=[positive_prompt, negative_prompt])
with gr.Accordion("Advanced Parameters: NCUT", open=False):
gr.Markdown("<a href='https://ncut-pytorch.readthedocs.io/en/latest/how_to_get_better_segmentation/' target='_blank'>Docs: How to Get Better Segmentation</a>")
affinity_focal_gamma_slider = gr.Slider(0.01, 1, step=0.01, label="NCUT: Affinity focal gamma", value=0.5, elem_id="affinity_focal_gamma", info="decrease for shaper segmentation")
num_sample_ncut_slider = gr.Slider(100, 50000, step=100, label="NCUT: num_sample", value=10000, elem_id="num_sample_ncut", info="Nyström approximation")
# sampling_method_dropdown = gr.Dropdown(["QuickFPS", "random"], label="NCUT: Sampling method", value="QuickFPS", elem_id="sampling_method", info="Nyström approximation")
sampling_method_dropdown = gr.Radio(["QuickFPS", "random"], label="NCUT: Sampling method", value="QuickFPS", elem_id="sampling_method")
# ncut_metric_dropdown = gr.Dropdown(["euclidean", "cosine"], label="NCUT: Distance metric", value="cosine", elem_id="ncut_metric")
ncut_metric_dropdown = gr.Radio(["euclidean", "cosine"], label="NCUT: Distance metric", value="cosine", elem_id="ncut_metric")
ncut_knn_slider = gr.Slider(1, 100, step=1, label="NCUT: KNN", value=10, elem_id="knn_ncut", info="Nyström approximation")
ncut_indirect_connection = gr.Checkbox(label="indirect_connection", value=False, elem_id="ncut_indirect_connection", info="TODO: Indirect connection is not implemented for directed NCUT", interactive=False)
ncut_make_orthogonal = gr.Checkbox(label="make_orthogonal", value=False, elem_id="ncut_make_orthogonal", info="Apply post-hoc eigenvectors orthogonalization")
with gr.Accordion("Advanced Parameters: Visualization", open=False):
# embedding_method_dropdown = gr.Dropdown(["tsne_3d", "umap_3d", "umap_sphere", "tsne_2d", "umap_2d"], label="Coloring method", value="tsne_3d", elem_id="embedding_method")
embedding_method_dropdown = gr.Radio(["tsne_3d", "umap_3d", "umap_sphere", "tsne_2d", "umap_2d"], label="Coloring method", value="tsne_3d", elem_id="embedding_method")
# embedding_metric_dropdown = gr.Dropdown(["euclidean", "cosine"], label="t-SNE/UMAP metric", value="euclidean", elem_id="embedding_metric")
embedding_metric_dropdown = gr.Radio(["euclidean", "cosine"], label="t-SNE/UMAP: metric", value="euclidean", elem_id="embedding_metric")
num_sample_tsne_slider = gr.Slider(100, 10000, step=100, label="t-SNE/UMAP: num_sample", value=300, elem_id="num_sample_tsne", info="Nyström approximation")
knn_tsne_slider = gr.Slider(1, 100, step=1, label="t-SNE/UMAP: KNN", value=10, elem_id="knn_tsne", info="Nyström approximation")
perplexity_slider = gr.Slider(10, 1000, step=10, label="t-SNE: perplexity", value=150, elem_id="perplexity")
n_neighbors_slider = gr.Slider(10, 1000, step=10, label="UMAP: n_neighbors", value=150, elem_id="n_neighbors")
min_dist_slider = gr.Slider(0.1, 1, step=0.1, label="UMAP: min_dist", value=0.1, elem_id="min_dist")
return [model_dropdown, layer_slider, node_type_dropdown, node_type_dropdown2, head_index_text, make_symmetric, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt]
def add_one_model(i_model=1):
with gr.Column(scale=5, min_width=200) as col:
gr.Markdown(f'### Output Images')
output_gallery = gr.Gallery(format='png', value=[], label="NCUT Embedding", show_label=True, elem_id=f"ncut{i_model}", columns=[3], rows=[1], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
submit_button = gr.Button("🔴 RUN", elem_id=f"submit_button{i_model}", variant='primary')
add_rotate_flip_buttons(output_gallery)
add_download_button(output_gallery, f"ncut_embed")
mlp_gallery = gr.Gallery(format='png', value=[], label="MLP color align", show_label=True, elem_id=f"mlp{i_model}", columns=[3], rows=[1], object_fit="contain", height="auto", show_fullscreen_button=True, interactive=False)
add_mlp_fitting_buttons(output_gallery, mlp_gallery)
add_download_button(mlp_gallery, f"mlp_color_align")
norm_gallery = gr.Gallery(value=[], label="Eigenvector Magnitude", show_label=True, elem_id=f"eig_norm{i_model}", columns=[3], rows=[1], object_fit="contain", height="auto", show_share_button=True, preview=False, interactive=False)
add_download_button(norm_gallery, f"eig_norm")
cluster_gallery = gr.Gallery(value=[], label="Clusters", show_label=True, elem_id=f"clusters{i_model}", columns=[2], rows=[4], object_fit="contain", height="auto", show_share_button=True, preview=False, interactive=False)
add_download_button(cluster_gallery, f"clusters")
[
model_dropdown, layer_slider, node_type_dropdown, node_type_dropdown2, head_index_text, make_symmetric, num_eig_slider,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider,
sampling_method_dropdown, ncut_metric_dropdown, positive_prompt, negative_prompt
] = make_parameters_section_2model()
# logging text box
logging_text = gr.Textbox("Logging information", label="Logging", elem_id="logging", type="text", placeholder="Logging information")
false_placeholder = gr.Checkbox(label="False", value=False, elem_id="false_placeholder", visible=False)
no_prompt = gr.Textbox("", label="", elem_id="empty_placeholder", type="text", placeholder="", visible=False)
false_placeholder = gr.Checkbox(label="False", value=False, elem_id="false_placeholder", visible=False)
submit_button.click(
partial(run_fn, n_ret=3, plot_clusters=True, alignedcut_eig_norm_plot=True, advanced=True, directed=True),
inputs=[
input_gallery, model_dropdown, layer_slider, num_eig_slider, node_type_dropdown,
positive_prompt, negative_prompt,
false_placeholder, no_prompt, no_prompt, no_prompt,
affinity_focal_gamma_slider, num_sample_ncut_slider, ncut_knn_slider, ncut_indirect_connection, ncut_make_orthogonal,
embedding_method_dropdown, embedding_metric_dropdown, num_sample_tsne_slider, knn_tsne_slider,
perplexity_slider, n_neighbors_slider, min_dist_slider, sampling_method_dropdown, ncut_metric_dropdown,
*[false_placeholder for _ in range(9)],
node_type_dropdown2, head_index_text, make_symmetric
],
outputs=[output_gallery, cluster_gallery, norm_gallery, logging_text]
)
output_gallery.change(lambda x: gr.update(value=x), inputs=[output_gallery], outputs=[mlp_gallery])
return output_gallery
galleries = []
with gr.Row():
with gr.Column(scale=5, min_width=200):
input_gallery, submit_button, clear_images_button, dataset_dropdown, num_images_slider, random_seed_slider, load_images_button = make_input_images_section(allow_download=True)
submit_button.visible = False
for i in range(3):
g = add_one_model()
galleries.append(g)
# Create rows and buttons in a loop
rows = []
buttons = []
for i in range(4):
row = gr.Row(visible=False)
rows.append(row)
with row:
for j in range(4):
with gr.Column(scale=5, min_width=200):
g = add_one_model()
galleries.append(g)
button = gr.Button("➕ Add Compare", elem_id=f"add_button_{i}", visible=False if i > 0 else True, scale=3)
buttons.append(button)
if i > 0:
# Reveal the current row and next button
buttons[i - 1].click(fn=lambda x: gr.update(visible=True), outputs=row)
buttons[i - 1].click(fn=lambda x: gr.update(visible=True), outputs=button)
# Hide the current button
buttons[i - 1].click(fn=lambda x: gr.update(visible=False), outputs=buttons[i - 1])
# Last button only reveals the last row and hides itself
buttons[-1].click(fn=lambda x: gr.update(visible=True), outputs=rows[-1])
buttons[-1].click(fn=lambda x: gr.update(visible=False), outputs=buttons[-1])
with gr.Tab('📄About'):
with gr.Column():
gr.Markdown("**This demo is for Python package `ncut-pytorch`, please visit the [Documentation](https://ncut-pytorch.readthedocs.io/)**")
gr.Markdown("**All the models and functions used for this demo are in the Python package `ncut-pytorch`**")
gr.Markdown("---")
gr.Markdown("---")
gr.Markdown("**Normalized Cuts**, aka. spectral clustering, is a graphical method to analyze data grouping in the affinity eigenvector space. It has been widely used for unsupervised segmentation in the 2000s.")
gr.Markdown("*Normalized Cuts and Image Segmentation, Jianbo Shi and Jitendra Malik, 2000*")
gr.Markdown("---")
gr.Markdown("**We have improved NCut, with some advanced features:**")
gr.Markdown("- **Nyström** Normalized Cut, is a new approximation algorithm developed for large-scale graph cuts, a large-graph of million nodes can be processed in under 10s (cpu) or 2s (gpu).")
gr.Markdown("- **spectral-tSNE** visualization, a new method to visualize the high-dimensional eigenvector space with 3D RGB cube. Color is aligned across images, color infers distance in representation.")
gr.Markdown("*paper in prep, Yang 2024*")
gr.Markdown("*AlignedCut: Visual Concepts Discovery on Brain-Guided Universal Feature Space, Huzheng Yang, James Gee\*, and Jianbo Shi\*, 2024*")
gr.Markdown("---")
gr.Markdown("---")
gr.Markdown('<p style="text-align: center;">We thank HuggingFace for hosting this demo.</p>')
# unlock the hidden tab
with gr.Row():
with gr.Column(scale=5):
gr.Markdown("")
with gr.Column(scale=5):
hidden_button = gr.Checkbox(label="🤗", value=False, elem_id="unlock_button", visible=True, interactive=True)
with gr.Column(scale=5):
gr.Markdown("")
n_smiles = gr.State(0)
unlock_value = 6
def update_smile(n_smiles):
n_smiles = n_smiles + 1
n_smiles = unlock_value if n_smiles > unlock_value else n_smiles
if n_smiles == unlock_value - 2:
gr.Info("click one more time to unlock", 2)
if n_smiles == unlock_value:
label = "🔓 unlocked"
return n_smiles, gr.update(label=label, value=True, interactive=False)
label = ["😊"] * n_smiles
label = "".join(label)
return n_smiles, gr.update(label=label, value=False)
def unlock_tabs_with_info(n_smiles):
if n_smiles == unlock_value:
gr.Info("🔓 unlocked tabs", 2)
return gr.update(visible=True)
return gr.update()
def unlock_tabs(n_smiles):
if n_smiles == unlock_value:
return gr.update(visible=True)
return gr.update()
hidden_button.change(update_smile, [n_smiles], [n_smiles, hidden_button])
hidden_button.change(unlock_tabs_with_info, n_smiles, tab_alignedcut_advanced)
hidden_button.change(unlock_tabs, n_smiles, tab_model_aligned_advanced)
hidden_button.change(unlock_tabs, n_smiles, tab_recursivecut_advanced)
hidden_button.change(unlock_tabs, n_smiles, tab_compare_models_advanced)
# with gr.Row():
# with gr.Column():
# gr.Markdown("##### This demo is for `ncut-pytorch`, [Documentation](https://ncut-pytorch.readthedocs.io/) ")
# with gr.Column():
# gr.Markdown("###### Running out of GPU Quota? Try [Demo](https://ncut-pytorch.readthedocs.io/en/latest/demo/) hosted at UPenn")
with gr.Row():
gr.Markdown("**This demo is for Python package `ncut-pytorch`, [Documentation](https://ncut-pytorch.readthedocs.io/)**")
# for local development
if os.path.exists("/hf_token.txt"):
os.environ["HF_ACCESS_TOKEN"] = open("/hf_token.txt").read().strip()
if DOWNLOAD_ALL_MODELS_DATASETS:
from ncut_pytorch.backbone import download_all_models
# t1 = threading.Thread(target=download_all_models).start()
# t1.join()
# t3 = threading.Thread(target=download_all_datasets).start()
# t3.join()
download_all_models()
download_all_datasets()
from ncut_pytorch.backbone_text import download_all_models
# t2 = threading.Thread(target=download_all_models).start()
# t2.join()
download_all_models()
demo.launch(share=True)
# # %%
# # debug
# # change working directory to "/"
# os.chdir("/")
# images = [(Image.open(image), None) for image in default_images]
# ret = run_fn(images, num_eig=30)
# # %%
# %%