Update app.py

app.py

@@ -2,7 +2,7 @@ import cv2, os, math
 import torch
 import random
 import numpy as np
-
+import json 
 import spaces
 
 import PIL
@@ -25,7 +25,7 @@ from insightface.app import FaceAnalysis
 from pipeline_controlnet_xs_sd_xl_instantid import StableDiffusionXLInstantIDXSPipeline, UNetControlNetXSModel
 
 from utils.controlnet_xs import ControlNetXSAdapter
-from style import styles
+
 
 
 import gradio as gr
@@ -39,7 +39,10 @@ hf_hub_download(repo_id="RED-AIGC/InstantID-XS", filename="image_proj.bin", loca
 MAX_SEED = np.iinfo(np.int32).max
 device = "cuda" if torch.cuda.is_available() else "cpu"
 weight_dtype = torch.float16 if str(device).__contains__("cuda") else torch.float32
-STYLE_NAMES = list(styles.keys())
+
+with open('./style.json') as f:
+    style_lib = json.load(f)
+STYLE_NAMES = list(style_lib.keys())
 DEFAULT_STYLE_NAME = "Ordinary"
 
 base_model = 'frankjoshua/realvisxlV40_v40Bakedvae'
@@ -117,29 +120,50 @@ def get_example():
             "./examples/1.jpg",
             None,
             "Ordinary",
-            "a woman",
-            "(lowres, low quality, worst quality:1.2), (text:1.2), watermark, (frame:1.2), deformed, ugly, deformed eyes, blur, out of focus, blurry, deformed cat, deformed, photo, anthropomorphic cat, monochrome, photo, pet collar, gun, weapon, blue, 3d, drones, drone, buildings in background, green",
+            ""
         ],
         [
             "./examples/1.jpg",
             "./examples/pose/pose1.jpg",
             "Hanfu",
-            "a woman",
-            "(lowres, low quality, worst quality:1.2), (text:1.2), watermark, (frame:1.2), deformed, ugly, deformed eyes, blur, out of focus, blurry, deformed cat, deformed, photo, anthropomorphic cat, monochrome, photo, pet collar, gun, weapon, blue, 3d, drones, drone, buildings in background, green",
+            ""
         ],
         [
             "./examples/2.jpg",
             "./examples/pose/pose2.png",
             "ZangZu",
-            "a woman",
-            "(lowres, low quality, worst quality:1.2), (text:1.2), watermark, (frame:1.2), deformed, ugly, deformed eyes, blur, out of focus, blurry, deformed cat, deformed, photo, anthropomorphic cat, monochrome, photo, pet collar, gun, weapon, blue, 3d, drones, drone, buildings in background, green",
+            ""
         ],
         [
             "./examples/3.png",
             "./examples/pose/pose3.png",
             "QingQiu",
-            "a woman",
-            "(lowres, low quality, worst quality:1.2), (text:1.2), watermark, (frame:1.2), deformed, ugly, deformed eyes, blur, out of focus, blurry, deformed cat, deformed, photo, anthropomorphic cat, monochrome, photo, pet collar, gun, weapon, blue, 3d, drones, drone, buildings in background, green",
+            "",
+        ],
+        [
+            "./examples/4.png",
+            "./examples/pose/pose2.png",
+            "(No style)",
+            "A man in suit",
+        ],
+
+        [
+            "./examples/5.jpeg",
+            "./examples/pose/pose3.png",
+            "(No style)",
+            "Girl in white wedding dress",
+        ],
+        [
+            "./examples/6.jpg",
+            "./examples/pose/pose4.jpeg",
+            "ZangZu",
+            "",
+        ],
+        [
+            "./examples/7.jpeg",
+            "./examples/pose/pose3.png",
+            "ZangZu",
+            "",
         ],
     ]
     return case
@@ -204,11 +228,12 @@ def resize_img(input_image,max_side=1280,min_side=1024,size=None,pad_to_max_side
         input_image = Image.fromarray(res)
     return input_image
 
-def apply_style(style_name: str, positive: str, negative: str = "") -> Tuple[str, str]:
-    p, n = styles.get(style_name, styles[DEFAULT_STYLE_NAME])
-    return p.replace("{prompt}", positive), n + ' ' + negative
+def apply_style(style_params, positive: str, negative: str = ""):
+    p = style_params["prompt"].replace("{prompt}", positive)
+    n = style_params["negative_prompt"] + ' ' + negative
+    return p, n
 
-def run_for_examples(face_file, pose_file, style, prompt, negative_prompt, ):
+def run_for_examples(face_file, pose_file, style, prompt, negative_prompt="", ):
     return generate_image(
         face_file,
         pose_file,
@@ -242,11 +267,6 @@ def generate_image(
     if face_image_path is None:
         raise gr.Error(f"Cannot find any input face image! Please upload the face image")
 
-    if prompt is None:
-        prompt = "a person"
-
-    # apply the style template
-    prompt, negative_prompt = apply_style(style_name, prompt, negative_prompt)
 
     face_image = load_image(face_image_path)
     face_image = resize_img(face_image, max_side=max_side)
@@ -267,6 +287,12 @@ def generate_image(
 
     face_emb = torch.from_numpy(face_info.normed_embedding)
     face_kps = draw_kps(convert_from_cv2_to_image(face_image_cv2), face_info["kps"])
+
+    style_params = style_lib[style_name][face_info["gender"]]
+    if prompt is None:
+        prompt = "a person"
+    prompt, negative_prompt = apply_style(style_params, prompt, negative_prompt)
+
     if pose_image_path is not None:
         pose_image = load_image(pose_image_path)
         pose_image = resize_img(pose_image, max_side=max_side)
@@ -342,6 +368,7 @@ with gr.Blocks(css=css) as demo:
 
             style = gr.Dropdown(
                 label="Style", 
+                info="If you want to generate images completely according to your own prompt, please choose '(No style)'",
                 choices=STYLE_NAMES, 
                 value=DEFAULT_STYLE_NAME
             )
@@ -432,7 +459,7 @@ with gr.Blocks(css=css) as demo:
 
     gr.Examples(
         examples=get_example(),
-        inputs=[face_file, pose_file, style, prompt, negative_prompt],
+        inputs=[face_file, pose_file, style, prompt],
         fn=run_for_examples,
         outputs=[gallery, usage_tips],
         cache_examples=True,