Upload 55 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+assets/compare_zoedepth.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,14 +1,114 @@
----
-title: Diffuse2PBR
-emoji: 🏆
-colorFrom: pink
-colorTo: yellow
-sdk: gradio
-sdk_version: 5.6.0
-app_file: app.py
-pinned: false
-license: cc-by-nc-2.0
-short_description: Convert Diffuse Textures to Height and Normal maps
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Depth Anything V2 for Metric Depth Estimation
+
+![teaser](./assets/compare_zoedepth.png)
+
+We here provide a simple codebase to fine-tune our Depth Anything V2 pre-trained encoder for metric depth estimation. Built on our powerful encoder, we use a simple DPT head to regress the depth. We fine-tune our pre-trained encoder on synthetic Hypersim / Virtual KITTI datasets for indoor / outdoor metric depth estimation, respectively.
+
+
+# Pre-trained Models
+
+We provide **six metric depth models** of three scales for indoor and outdoor scenes, respectively.
+
+| Base Model | Params | Indoor (Hypersim) | Outdoor (Virtual KITTI 2) |
+|:-|-:|:-:|:-:|
+| Depth-Anything-V2-Small | 24.8M | [Download](https://huggingface.co/depth-anything/Depth-Anything-V2-Metric-Hypersim-Small/resolve/main/depth_anything_v2_metric_hypersim_vits.pth?download=true) | [Download](https://huggingface.co/depth-anything/Depth-Anything-V2-Metric-VKITTI-Small/resolve/main/depth_anything_v2_metric_vkitti_vits.pth?download=true) |
+| Depth-Anything-V2-Base | 97.5M | [Download](https://huggingface.co/depth-anything/Depth-Anything-V2-Metric-Hypersim-Base/resolve/main/depth_anything_v2_metric_hypersim_vitb.pth?download=true) | [Download](https://huggingface.co/depth-anything/Depth-Anything-V2-Metric-VKITTI-Base/resolve/main/depth_anything_v2_metric_vkitti_vitb.pth?download=true) |
+| Depth-Anything-V2-Large | 335.3M | [Download](https://huggingface.co/depth-anything/Depth-Anything-V2-Metric-Hypersim-Large/resolve/main/depth_anything_v2_metric_hypersim_vitl.pth?download=true) | [Download](https://huggingface.co/depth-anything/Depth-Anything-V2-Metric-VKITTI-Large/resolve/main/depth_anything_v2_metric_vkitti_vitl.pth?download=true) |
+
+*We recommend to first try our larger models (if computational cost is affordable) and the indoor version.*
+
+## Usage
+
+### Prepraration
+
+```bash
+git clone https://github.com/DepthAnything/Depth-Anything-V2
+cd Depth-Anything-V2/metric_depth
+pip install -r requirements.txt
+```
+
+Download the checkpoints listed [here](#pre-trained-models) and put them under the `checkpoints` directory.
+
+### Use our models
+```python
+import cv2
+import torch
+
+from depth_anything_v2.dpt import DepthAnythingV2
+
+model_configs = {
+    'vits': {'encoder': 'vits', 'features': 64, 'out_channels': [48, 96, 192, 384]},
+    'vitb': {'encoder': 'vitb', 'features': 128, 'out_channels': [96, 192, 384, 768]},
+    'vitl': {'encoder': 'vitl', 'features': 256, 'out_channels': [256, 512, 1024, 1024]}
+}
+
+encoder = 'vitl' # or 'vits', 'vitb'
+dataset = 'hypersim' # 'hypersim' for indoor model, 'vkitti' for outdoor model
+max_depth = 20 # 20 for indoor model, 80 for outdoor model
+
+model = DepthAnythingV2(**{**model_configs[encoder], 'max_depth': max_depth})
+model.load_state_dict(torch.load(f'checkpoints/depth_anything_v2_metric_{dataset}_{encoder}.pth', map_location='cpu'))
+model.eval()
+
+raw_img = cv2.imread('your/image/path')
+depth = model.infer_image(raw_img) # HxW depth map in meters in numpy
+```
+
+### Running script on images
+
+Here, we take the `vitl` encoder as an example. You can also use `vitb` or `vits` encoders.
+
+```bash
+# indoor scenes
+python run.py \
+  --encoder vitl \
+  --load-from checkpoints/depth_anything_v2_metric_hypersim_vitl.pth \
+  --max-depth 20 \
+  --img-path <path> --outdir <outdir> [--input-size <size>] [--save-numpy]
+
+# outdoor scenes
+python run.py \
+  --encoder vitl \
+  --load-from checkpoints/depth_anything_v2_metric_vkitti_vitl.pth \
+  --max-depth 80 \
+  --img-path <path> --outdir <outdir> [--input-size <size>] [--save-numpy]
+```
+
+### Project 2D images to point clouds:
+
+```bash
+python depth_to_pointcloud.py \
+  --encoder vitl \
+  --load-from checkpoints/depth_anything_v2_metric_hypersim_vitl.pth \
+  --max-depth 20 \
+  --img-path <path> --outdir <outdir>
+```
+
+### Reproduce training
+
+Please first prepare the [Hypersim](https://github.com/apple/ml-hypersim) and [Virtual KITTI 2](https://europe.naverlabs.com/research/computer-vision/proxy-virtual-worlds-vkitti-2/) datasets. Then:
+
+```bash
+bash dist_train.sh
+```
+
+
+## Citation
+
+If you find this project useful, please consider citing:
+
+```bibtex
+@article{depth_anything_v2,
+  title={Depth Anything V2},
+  author={Yang, Lihe and Kang, Bingyi and Huang, Zilong and Zhao, Zhen and Xu, Xiaogang and Feng, Jiashi and Zhao, Hengshuang},
+  journal={arXiv:2406.09414},
+  year={2024}
+}
+
+@inproceedings{depth_anything_v1,
+  title={Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data}, 
+  author={Yang, Lihe and Kang, Bingyi and Huang, Zilong and Xu, Xiaogang and Feng, Jiashi and Zhao, Hengshuang},
+  booktitle={CVPR},
+  year={2024}
+}
+```

depth_anything_v2/dinov2.py

@@ -0,0 +1,415 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+#
+# This source code is licensed under the Apache License, Version 2.0
+# found in the LICENSE file in the root directory of this source tree.
+
+# References:
+#   https://github.com/facebookresearch/dino/blob/main/vision_transformer.py
+#   https://github.com/rwightman/pytorch-image-models/tree/master/timm/models/vision_transformer.py
+
+from functools import partial
+import math
+import logging
+from typing import Sequence, Tuple, Union, Callable
+
+import torch
+import torch.nn as nn
+import torch.utils.checkpoint
+from torch.nn.init import trunc_normal_
+
+from .dinov2_layers import Mlp, PatchEmbed, SwiGLUFFNFused, MemEffAttention, NestedTensorBlock as Block
+
+
+logger = logging.getLogger("dinov2")
+
+
+def named_apply(fn: Callable, module: nn.Module, name="", depth_first=True, include_root=False) -> nn.Module:
+    if not depth_first and include_root:
+        fn(module=module, name=name)
+    for child_name, child_module in module.named_children():
+        child_name = ".".join((name, child_name)) if name else child_name
+        named_apply(fn=fn, module=child_module, name=child_name, depth_first=depth_first, include_root=True)
+    if depth_first and include_root:
+        fn(module=module, name=name)
+    return module
+
+
+class BlockChunk(nn.ModuleList):
+    def forward(self, x):
+        for b in self:
+            x = b(x)
+        return x
+
+
+class DinoVisionTransformer(nn.Module):
+    def __init__(
+        self,
+        img_size=224,
+        patch_size=16,
+        in_chans=3,
+        embed_dim=768,
+        depth=12,
+        num_heads=12,
+        mlp_ratio=4.0,
+        qkv_bias=True,
+        ffn_bias=True,
+        proj_bias=True,
+        drop_path_rate=0.0,
+        drop_path_uniform=False,
+        init_values=None,  # for layerscale: None or 0 => no layerscale
+        embed_layer=PatchEmbed,
+        act_layer=nn.GELU,
+        block_fn=Block,
+        ffn_layer="mlp",
+        block_chunks=1,
+        num_register_tokens=0,
+        interpolate_antialias=False,
+        interpolate_offset=0.1,
+    ):
+        """
+        Args:
+            img_size (int, tuple): input image size
+            patch_size (int, tuple): patch size
+            in_chans (int): number of input channels
+            embed_dim (int): embedding dimension
+            depth (int): depth of transformer
+            num_heads (int): number of attention heads
+            mlp_ratio (int): ratio of mlp hidden dim to embedding dim
+            qkv_bias (bool): enable bias for qkv if True
+            proj_bias (bool): enable bias for proj in attn if True
+            ffn_bias (bool): enable bias for ffn if True
+            drop_path_rate (float): stochastic depth rate
+            drop_path_uniform (bool): apply uniform drop rate across blocks
+            weight_init (str): weight init scheme
+            init_values (float): layer-scale init values
+            embed_layer (nn.Module): patch embedding layer
+            act_layer (nn.Module): MLP activation layer
+            block_fn (nn.Module): transformer block class
+            ffn_layer (str): "mlp", "swiglu", "swiglufused" or "identity"
+            block_chunks: (int) split block sequence into block_chunks units for FSDP wrap
+            num_register_tokens: (int) number of extra cls tokens (so-called "registers")
+            interpolate_antialias: (str) flag to apply anti-aliasing when interpolating positional embeddings
+            interpolate_offset: (float) work-around offset to apply when interpolating positional embeddings
+        """
+        super().__init__()
+        norm_layer = partial(nn.LayerNorm, eps=1e-6)
+
+        self.num_features = self.embed_dim = embed_dim  # num_features for consistency with other models
+        self.num_tokens = 1
+        self.n_blocks = depth
+        self.num_heads = num_heads
+        self.patch_size = patch_size
+        self.num_register_tokens = num_register_tokens
+        self.interpolate_antialias = interpolate_antialias
+        self.interpolate_offset = interpolate_offset
+
+        self.patch_embed = embed_layer(img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim)
+        num_patches = self.patch_embed.num_patches
+
+        self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
+        self.pos_embed = nn.Parameter(torch.zeros(1, num_patches + self.num_tokens, embed_dim))
+        assert num_register_tokens >= 0
+        self.register_tokens = (
+            nn.Parameter(torch.zeros(1, num_register_tokens, embed_dim)) if num_register_tokens else None
+        )
+
+        if drop_path_uniform is True:
+            dpr = [drop_path_rate] * depth
+        else:
+            dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)]  # stochastic depth decay rule
+
+        if ffn_layer == "mlp":
+            logger.info("using MLP layer as FFN")
+            ffn_layer = Mlp
+        elif ffn_layer == "swiglufused" or ffn_layer == "swiglu":
+            logger.info("using SwiGLU layer as FFN")
+            ffn_layer = SwiGLUFFNFused
+        elif ffn_layer == "identity":
+            logger.info("using Identity layer as FFN")
+
+            def f(*args, **kwargs):
+                return nn.Identity()
+
+            ffn_layer = f
+        else:
+            raise NotImplementedError
+
+        blocks_list = [
+            block_fn(
+                dim=embed_dim,
+                num_heads=num_heads,
+                mlp_ratio=mlp_ratio,
+                qkv_bias=qkv_bias,
+                proj_bias=proj_bias,
+                ffn_bias=ffn_bias,
+                drop_path=dpr[i],
+                norm_layer=norm_layer,
+                act_layer=act_layer,
+                ffn_layer=ffn_layer,
+                init_values=init_values,
+            )
+            for i in range(depth)
+        ]
+        if block_chunks > 0:
+            self.chunked_blocks = True
+            chunked_blocks = []
+            chunksize = depth // block_chunks
+            for i in range(0, depth, chunksize):
+                # this is to keep the block index consistent if we chunk the block list
+                chunked_blocks.append([nn.Identity()] * i + blocks_list[i : i + chunksize])
+            self.blocks = nn.ModuleList([BlockChunk(p) for p in chunked_blocks])
+        else:
+            self.chunked_blocks = False
+            self.blocks = nn.ModuleList(blocks_list)
+
+        self.norm = norm_layer(embed_dim)
+        self.head = nn.Identity()
+
+        self.mask_token = nn.Parameter(torch.zeros(1, embed_dim))
+
+        self.init_weights()
+
+    def init_weights(self):
+        trunc_normal_(self.pos_embed, std=0.02)
+        nn.init.normal_(self.cls_token, std=1e-6)
+        if self.register_tokens is not None:
+            nn.init.normal_(self.register_tokens, std=1e-6)
+        named_apply(init_weights_vit_timm, self)
+
+    def interpolate_pos_encoding(self, x, w, h):
+        previous_dtype = x.dtype
+        npatch = x.shape[1] - 1
+        N = self.pos_embed.shape[1] - 1
+        if npatch == N and w == h:
+            return self.pos_embed
+        pos_embed = self.pos_embed.float()
+        class_pos_embed = pos_embed[:, 0]
+        patch_pos_embed = pos_embed[:, 1:]
+        dim = x.shape[-1]
+        w0 = w // self.patch_size
+        h0 = h // self.patch_size
+        # we add a small number to avoid floating point error in the interpolation
+        # see discussion at https://github.com/facebookresearch/dino/issues/8
+        # DINOv2 with register modify the interpolate_offset from 0.1 to 0.0
+        w0, h0 = w0 + self.interpolate_offset, h0 + self.interpolate_offset
+        # w0, h0 = w0 + 0.1, h0 + 0.1
+        
+        sqrt_N = math.sqrt(N)
+        sx, sy = float(w0) / sqrt_N, float(h0) / sqrt_N
+        patch_pos_embed = nn.functional.interpolate(
+            patch_pos_embed.reshape(1, int(sqrt_N), int(sqrt_N), dim).permute(0, 3, 1, 2),
+            scale_factor=(sx, sy),
+            # (int(w0), int(h0)), # to solve the upsampling shape issue
+            mode="bicubic",
+            antialias=self.interpolate_antialias
+        )
+        
+        assert int(w0) == patch_pos_embed.shape[-2]
+        assert int(h0) == patch_pos_embed.shape[-1]
+        patch_pos_embed = patch_pos_embed.permute(0, 2, 3, 1).view(1, -1, dim)
+        return torch.cat((class_pos_embed.unsqueeze(0), patch_pos_embed), dim=1).to(previous_dtype)
+
+    def prepare_tokens_with_masks(self, x, masks=None):
+        B, nc, w, h = x.shape
+        x = self.patch_embed(x)
+        if masks is not None:
+            x = torch.where(masks.unsqueeze(-1), self.mask_token.to(x.dtype).unsqueeze(0), x)
+
+        x = torch.cat((self.cls_token.expand(x.shape[0], -1, -1), x), dim=1)
+        x = x + self.interpolate_pos_encoding(x, w, h)
+
+        if self.register_tokens is not None:
+            x = torch.cat(
+                (
+                    x[:, :1],
+                    self.register_tokens.expand(x.shape[0], -1, -1),
+                    x[:, 1:],
+                ),
+                dim=1,
+            )
+
+        return x
+
+    def forward_features_list(self, x_list, masks_list):
+        x = [self.prepare_tokens_with_masks(x, masks) for x, masks in zip(x_list, masks_list)]
+        for blk in self.blocks:
+            x = blk(x)
+
+        all_x = x
+        output = []
+        for x, masks in zip(all_x, masks_list):
+            x_norm = self.norm(x)
+            output.append(
+                {
+                    "x_norm_clstoken": x_norm[:, 0],
+                    "x_norm_regtokens": x_norm[:, 1 : self.num_register_tokens + 1],
+                    "x_norm_patchtokens": x_norm[:, self.num_register_tokens + 1 :],
+                    "x_prenorm": x,
+                    "masks": masks,
+                }
+            )
+        return output
+
+    def forward_features(self, x, masks=None):
+        if isinstance(x, list):
+            return self.forward_features_list(x, masks)
+
+        x = self.prepare_tokens_with_masks(x, masks)
+
+        for blk in self.blocks:
+            x = blk(x)
+
+        x_norm = self.norm(x)
+        return {
+            "x_norm_clstoken": x_norm[:, 0],
+            "x_norm_regtokens": x_norm[:, 1 : self.num_register_tokens + 1],
+            "x_norm_patchtokens": x_norm[:, self.num_register_tokens + 1 :],
+            "x_prenorm": x,
+            "masks": masks,
+        }
+
+    def _get_intermediate_layers_not_chunked(self, x, n=1):
+        x = self.prepare_tokens_with_masks(x)
+        # If n is an int, take the n last blocks. If it's a list, take them
+        output, total_block_len = [], len(self.blocks)
+        blocks_to_take = range(total_block_len - n, total_block_len) if isinstance(n, int) else n
+        for i, blk in enumerate(self.blocks):
+            x = blk(x)
+            if i in blocks_to_take:
+                output.append(x)
+        assert len(output) == len(blocks_to_take), f"only {len(output)} / {len(blocks_to_take)} blocks found"
+        return output
+
+    def _get_intermediate_layers_chunked(self, x, n=1):
+        x = self.prepare_tokens_with_masks(x)
+        output, i, total_block_len = [], 0, len(self.blocks[-1])
+        # If n is an int, take the n last blocks. If it's a list, take them
+        blocks_to_take = range(total_block_len - n, total_block_len) if isinstance(n, int) else n
+        for block_chunk in self.blocks:
+            for blk in block_chunk[i:]:  # Passing the nn.Identity()
+                x = blk(x)
+                if i in blocks_to_take:
+                    output.append(x)
+                i += 1
+        assert len(output) == len(blocks_to_take), f"only {len(output)} / {len(blocks_to_take)} blocks found"
+        return output
+
+    def get_intermediate_layers(
+        self,
+        x: torch.Tensor,
+        n: Union[int, Sequence] = 1,  # Layers or n last layers to take
+        reshape: bool = False,
+        return_class_token: bool = False,
+        norm=True
+    ) -> Tuple[Union[torch.Tensor, Tuple[torch.Tensor]]]:
+        if self.chunked_blocks:
+            outputs = self._get_intermediate_layers_chunked(x, n)
+        else:
+            outputs = self._get_intermediate_layers_not_chunked(x, n)
+        if norm:
+            outputs = [self.norm(out) for out in outputs]
+        class_tokens = [out[:, 0] for out in outputs]
+        outputs = [out[:, 1 + self.num_register_tokens:] for out in outputs]
+        if reshape:
+            B, _, w, h = x.shape
+            outputs = [
+                out.reshape(B, w // self.patch_size, h // self.patch_size, -1).permute(0, 3, 1, 2).contiguous()
+                for out in outputs
+            ]
+        if return_class_token:
+            return tuple(zip(outputs, class_tokens))
+        return tuple(outputs)
+
+    def forward(self, *args, is_training=False, **kwargs):
+        ret = self.forward_features(*args, **kwargs)
+        if is_training:
+            return ret
+        else:
+            return self.head(ret["x_norm_clstoken"])
+
+
+def init_weights_vit_timm(module: nn.Module, name: str = ""):
+    """ViT weight initialization, original timm impl (for reproducibility)"""
+    if isinstance(module, nn.Linear):
+        trunc_normal_(module.weight, std=0.02)
+        if module.bias is not None:
+            nn.init.zeros_(module.bias)
+
+
+def vit_small(patch_size=16, num_register_tokens=0, **kwargs):
+    model = DinoVisionTransformer(
+        patch_size=patch_size,
+        embed_dim=384,
+        depth=12,
+        num_heads=6,
+        mlp_ratio=4,
+        block_fn=partial(Block, attn_class=MemEffAttention),
+        num_register_tokens=num_register_tokens,
+        **kwargs,
+    )
+    return model
+
+
+def vit_base(patch_size=16, num_register_tokens=0, **kwargs):
+    model = DinoVisionTransformer(
+        patch_size=patch_size,
+        embed_dim=768,
+        depth=12,
+        num_heads=12,
+        mlp_ratio=4,
+        block_fn=partial(Block, attn_class=MemEffAttention),
+        num_register_tokens=num_register_tokens,
+        **kwargs,
+    )
+    return model
+
+
+def vit_large(patch_size=16, num_register_tokens=0, **kwargs):
+    model = DinoVisionTransformer(
+        patch_size=patch_size,
+        embed_dim=1024,
+        depth=24,
+        num_heads=16,
+        mlp_ratio=4,
+        block_fn=partial(Block, attn_class=MemEffAttention),
+        num_register_tokens=num_register_tokens,
+        **kwargs,
+    )
+    return model
+
+
+def vit_giant2(patch_size=16, num_register_tokens=0, **kwargs):
+    """
+    Close to ViT-giant, with embed-dim 1536 and 24 heads => embed-dim per head 64
+    """
+    model = DinoVisionTransformer(
+        patch_size=patch_size,
+        embed_dim=1536,
+        depth=40,
+        num_heads=24,
+        mlp_ratio=4,
+        block_fn=partial(Block, attn_class=MemEffAttention),
+        num_register_tokens=num_register_tokens,
+        **kwargs,
+    )
+    return model
+
+
+def DINOv2(model_name):
+    model_zoo = {
+        "vits": vit_small, 
+        "vitb": vit_base, 
+        "vitl": vit_large, 
+        "vitg": vit_giant2
+    }
+    
+    return model_zoo[model_name](
+        img_size=518,
+        patch_size=14,
+        init_values=1.0,
+        ffn_layer="mlp" if model_name != "vitg" else "swiglufused",
+        block_chunks=0,
+        num_register_tokens=0,
+        interpolate_antialias=False,
+        interpolate_offset=0.1
+    )

depth_anything_v2/dinov2_layers/__init__.py

@@ -0,0 +1,11 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from .mlp import Mlp
+from .patch_embed import PatchEmbed
+from .swiglu_ffn import SwiGLUFFN, SwiGLUFFNFused
+from .block import NestedTensorBlock
+from .attention import MemEffAttention

depth_anything_v2/dinov2_layers/attention.py

@@ -0,0 +1,83 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+# References:
+#   https://github.com/facebookresearch/dino/blob/master/vision_transformer.py
+#   https://github.com/rwightman/pytorch-image-models/tree/master/timm/models/vision_transformer.py
+
+import logging
+
+from torch import Tensor
+from torch import nn
+
+
+logger = logging.getLogger("dinov2")
+
+
+try:
+    from xformers.ops import memory_efficient_attention, unbind, fmha
+
+    XFORMERS_AVAILABLE = True
+except ImportError:
+    logger.warning("xFormers not available")
+    XFORMERS_AVAILABLE = False
+
+
+class Attention(nn.Module):
+    def __init__(
+        self,
+        dim: int,
+        num_heads: int = 8,
+        qkv_bias: bool = False,
+        proj_bias: bool = True,
+        attn_drop: float = 0.0,
+        proj_drop: float = 0.0,
+    ) -> None:
+        super().__init__()
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+        self.scale = head_dim**-0.5
+
+        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
+        self.attn_drop = nn.Dropout(attn_drop)
+        self.proj = nn.Linear(dim, dim, bias=proj_bias)
+        self.proj_drop = nn.Dropout(proj_drop)
+
+    def forward(self, x: Tensor) -> Tensor:
+        B, N, C = x.shape
+        qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
+
+        q, k, v = qkv[0] * self.scale, qkv[1], qkv[2]
+        attn = q @ k.transpose(-2, -1)
+
+        attn = attn.softmax(dim=-1)
+        attn = self.attn_drop(attn)
+
+        x = (attn @ v).transpose(1, 2).reshape(B, N, C)
+        x = self.proj(x)
+        x = self.proj_drop(x)
+        return x
+
+
+class MemEffAttention(Attention):
+    def forward(self, x: Tensor, attn_bias=None) -> Tensor:
+        if not XFORMERS_AVAILABLE:
+            assert attn_bias is None, "xFormers is required for nested tensors usage"
+            return super().forward(x)
+
+        B, N, C = x.shape
+        qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads)
+
+        q, k, v = unbind(qkv, 2)
+
+        x = memory_efficient_attention(q, k, v, attn_bias=attn_bias)
+        x = x.reshape([B, N, C])
+
+        x = self.proj(x)
+        x = self.proj_drop(x)
+        return x
+
+        

depth_anything_v2/dinov2_layers/block.py

@@ -0,0 +1,252 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+# References:
+#   https://github.com/facebookresearch/dino/blob/master/vision_transformer.py
+#   https://github.com/rwightman/pytorch-image-models/tree/master/timm/layers/patch_embed.py
+
+import logging
+from typing import Callable, List, Any, Tuple, Dict
+
+import torch
+from torch import nn, Tensor
+
+from .attention import Attention, MemEffAttention
+from .drop_path import DropPath
+from .layer_scale import LayerScale
+from .mlp import Mlp
+
+
+logger = logging.getLogger("dinov2")
+
+
+try:
+    from xformers.ops import fmha
+    from xformers.ops import scaled_index_add, index_select_cat
+
+    XFORMERS_AVAILABLE = True
+except ImportError:
+    logger.warning("xFormers not available")
+    XFORMERS_AVAILABLE = False
+
+
+class Block(nn.Module):
+    def __init__(
+        self,
+        dim: int,
+        num_heads: int,
+        mlp_ratio: float = 4.0,
+        qkv_bias: bool = False,
+        proj_bias: bool = True,
+        ffn_bias: bool = True,
+        drop: float = 0.0,
+        attn_drop: float = 0.0,
+        init_values=None,
+        drop_path: float = 0.0,
+        act_layer: Callable[..., nn.Module] = nn.GELU,
+        norm_layer: Callable[..., nn.Module] = nn.LayerNorm,
+        attn_class: Callable[..., nn.Module] = Attention,
+        ffn_layer: Callable[..., nn.Module] = Mlp,
+    ) -> None:
+        super().__init__()
+        # print(f"biases: qkv: {qkv_bias}, proj: {proj_bias}, ffn: {ffn_bias}")
+        self.norm1 = norm_layer(dim)
+        self.attn = attn_class(
+            dim,
+            num_heads=num_heads,
+            qkv_bias=qkv_bias,
+            proj_bias=proj_bias,
+            attn_drop=attn_drop,
+            proj_drop=drop,
+        )
+        self.ls1 = LayerScale(dim, init_values=init_values) if init_values else nn.Identity()
+        self.drop_path1 = DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
+
+        self.norm2 = norm_layer(dim)
+        mlp_hidden_dim = int(dim * mlp_ratio)
+        self.mlp = ffn_layer(
+            in_features=dim,
+            hidden_features=mlp_hidden_dim,
+            act_layer=act_layer,
+            drop=drop,
+            bias=ffn_bias,
+        )
+        self.ls2 = LayerScale(dim, init_values=init_values) if init_values else nn.Identity()
+        self.drop_path2 = DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
+
+        self.sample_drop_ratio = drop_path
+
+    def forward(self, x: Tensor) -> Tensor:
+        def attn_residual_func(x: Tensor) -> Tensor:
+            return self.ls1(self.attn(self.norm1(x)))
+
+        def ffn_residual_func(x: Tensor) -> Tensor:
+            return self.ls2(self.mlp(self.norm2(x)))
+
+        if self.training and self.sample_drop_ratio > 0.1:
+            # the overhead is compensated only for a drop path rate larger than 0.1
+            x = drop_add_residual_stochastic_depth(
+                x,
+                residual_func=attn_residual_func,
+                sample_drop_ratio=self.sample_drop_ratio,
+            )
+            x = drop_add_residual_stochastic_depth(
+                x,
+                residual_func=ffn_residual_func,
+                sample_drop_ratio=self.sample_drop_ratio,
+            )
+        elif self.training and self.sample_drop_ratio > 0.0:
+            x = x + self.drop_path1(attn_residual_func(x))
+            x = x + self.drop_path1(ffn_residual_func(x))  # FIXME: drop_path2
+        else:
+            x = x + attn_residual_func(x)
+            x = x + ffn_residual_func(x)
+        return x
+
+
+def drop_add_residual_stochastic_depth(
+    x: Tensor,
+    residual_func: Callable[[Tensor], Tensor],
+    sample_drop_ratio: float = 0.0,
+) -> Tensor:
+    # 1) extract subset using permutation
+    b, n, d = x.shape
+    sample_subset_size = max(int(b * (1 - sample_drop_ratio)), 1)
+    brange = (torch.randperm(b, device=x.device))[:sample_subset_size]
+    x_subset = x[brange]
+
+    # 2) apply residual_func to get residual
+    residual = residual_func(x_subset)
+
+    x_flat = x.flatten(1)
+    residual = residual.flatten(1)
+
+    residual_scale_factor = b / sample_subset_size
+
+    # 3) add the residual
+    x_plus_residual = torch.index_add(x_flat, 0, brange, residual.to(dtype=x.dtype), alpha=residual_scale_factor)
+    return x_plus_residual.view_as(x)
+
+
+def get_branges_scales(x, sample_drop_ratio=0.0):
+    b, n, d = x.shape
+    sample_subset_size = max(int(b * (1 - sample_drop_ratio)), 1)
+    brange = (torch.randperm(b, device=x.device))[:sample_subset_size]
+    residual_scale_factor = b / sample_subset_size
+    return brange, residual_scale_factor
+
+
+def add_residual(x, brange, residual, residual_scale_factor, scaling_vector=None):
+    if scaling_vector is None:
+        x_flat = x.flatten(1)
+        residual = residual.flatten(1)
+        x_plus_residual = torch.index_add(x_flat, 0, brange, residual.to(dtype=x.dtype), alpha=residual_scale_factor)
+    else:
+        x_plus_residual = scaled_index_add(
+            x, brange, residual.to(dtype=x.dtype), scaling=scaling_vector, alpha=residual_scale_factor
+        )
+    return x_plus_residual
+
+
+attn_bias_cache: Dict[Tuple, Any] = {}
+
+
+def get_attn_bias_and_cat(x_list, branges=None):
+    """
+    this will perform the index select, cat the tensors, and provide the attn_bias from cache
+    """
+    batch_sizes = [b.shape[0] for b in branges] if branges is not None else [x.shape[0] for x in x_list]
+    all_shapes = tuple((b, x.shape[1]) for b, x in zip(batch_sizes, x_list))
+    if all_shapes not in attn_bias_cache.keys():
+        seqlens = []
+        for b, x in zip(batch_sizes, x_list):
+            for _ in range(b):
+                seqlens.append(x.shape[1])
+        attn_bias = fmha.BlockDiagonalMask.from_seqlens(seqlens)
+        attn_bias._batch_sizes = batch_sizes
+        attn_bias_cache[all_shapes] = attn_bias
+
+    if branges is not None:
+        cat_tensors = index_select_cat([x.flatten(1) for x in x_list], branges).view(1, -1, x_list[0].shape[-1])
+    else:
+        tensors_bs1 = tuple(x.reshape([1, -1, *x.shape[2:]]) for x in x_list)
+        cat_tensors = torch.cat(tensors_bs1, dim=1)
+
+    return attn_bias_cache[all_shapes], cat_tensors
+
+
+def drop_add_residual_stochastic_depth_list(
+    x_list: List[Tensor],
+    residual_func: Callable[[Tensor, Any], Tensor],
+    sample_drop_ratio: float = 0.0,
+    scaling_vector=None,
+) -> Tensor:
+    # 1) generate random set of indices for dropping samples in the batch
+    branges_scales = [get_branges_scales(x, sample_drop_ratio=sample_drop_ratio) for x in x_list]
+    branges = [s[0] for s in branges_scales]
+    residual_scale_factors = [s[1] for s in branges_scales]
+
+    # 2) get attention bias and index+concat the tensors
+    attn_bias, x_cat = get_attn_bias_and_cat(x_list, branges)
+
+    # 3) apply residual_func to get residual, and split the result
+    residual_list = attn_bias.split(residual_func(x_cat, attn_bias=attn_bias))  # type: ignore
+
+    outputs = []
+    for x, brange, residual, residual_scale_factor in zip(x_list, branges, residual_list, residual_scale_factors):
+        outputs.append(add_residual(x, brange, residual, residual_scale_factor, scaling_vector).view_as(x))
+    return outputs
+
+
+class NestedTensorBlock(Block):
+    def forward_nested(self, x_list: List[Tensor]) -> List[Tensor]:
+        """
+        x_list contains a list of tensors to nest together and run
+        """
+        assert isinstance(self.attn, MemEffAttention)
+
+        if self.training and self.sample_drop_ratio > 0.0:
+
+            def attn_residual_func(x: Tensor, attn_bias=None) -> Tensor:
+                return self.attn(self.norm1(x), attn_bias=attn_bias)
+
+            def ffn_residual_func(x: Tensor, attn_bias=None) -> Tensor:
+                return self.mlp(self.norm2(x))
+
+            x_list = drop_add_residual_stochastic_depth_list(
+                x_list,
+                residual_func=attn_residual_func,
+                sample_drop_ratio=self.sample_drop_ratio,
+                scaling_vector=self.ls1.gamma if isinstance(self.ls1, LayerScale) else None,
+            )
+            x_list = drop_add_residual_stochastic_depth_list(
+                x_list,
+                residual_func=ffn_residual_func,
+                sample_drop_ratio=self.sample_drop_ratio,
+                scaling_vector=self.ls2.gamma if isinstance(self.ls1, LayerScale) else None,
+            )
+            return x_list
+        else:
+
+            def attn_residual_func(x: Tensor, attn_bias=None) -> Tensor:
+                return self.ls1(self.attn(self.norm1(x), attn_bias=attn_bias))
+
+            def ffn_residual_func(x: Tensor, attn_bias=None) -> Tensor:
+                return self.ls2(self.mlp(self.norm2(x)))
+
+            attn_bias, x = get_attn_bias_and_cat(x_list)
+            x = x + attn_residual_func(x, attn_bias=attn_bias)
+            x = x + ffn_residual_func(x)
+            return attn_bias.split(x)
+
+    def forward(self, x_or_x_list):
+        if isinstance(x_or_x_list, Tensor):
+            return super().forward(x_or_x_list)
+        elif isinstance(x_or_x_list, list):
+            assert XFORMERS_AVAILABLE, "Please install xFormers for nested tensors usage"
+            return self.forward_nested(x_or_x_list)
+        else:
+            raise AssertionError

depth_anything_v2/dinov2_layers/drop_path.py

@@ -0,0 +1,35 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+# References:
+#   https://github.com/facebookresearch/dino/blob/master/vision_transformer.py
+#   https://github.com/rwightman/pytorch-image-models/tree/master/timm/layers/drop.py
+
+
+from torch import nn
+
+
+def drop_path(x, drop_prob: float = 0.0, training: bool = False):
+    if drop_prob == 0.0 or not training:
+        return x
+    keep_prob = 1 - drop_prob
+    shape = (x.shape[0],) + (1,) * (x.ndim - 1)  # work with diff dim tensors, not just 2D ConvNets
+    random_tensor = x.new_empty(shape).bernoulli_(keep_prob)
+    if keep_prob > 0.0:
+        random_tensor.div_(keep_prob)
+    output = x * random_tensor
+    return output
+
+
+class DropPath(nn.Module):
+    """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks)."""
+
+    def __init__(self, drop_prob=None):
+        super(DropPath, self).__init__()
+        self.drop_prob = drop_prob
+
+    def forward(self, x):
+        return drop_path(x, self.drop_prob, self.training)

depth_anything_v2/dinov2_layers/layer_scale.py

@@ -0,0 +1,28 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+# Modified from: https://github.com/huggingface/pytorch-image-models/blob/main/timm/models/vision_transformer.py#L103-L110
+
+from typing import Union
+
+import torch
+from torch import Tensor
+from torch import nn
+
+
+class LayerScale(nn.Module):
+    def __init__(
+        self,
+        dim: int,
+        init_values: Union[float, Tensor] = 1e-5,
+        inplace: bool = False,
+    ) -> None:
+        super().__init__()
+        self.inplace = inplace
+        self.gamma = nn.Parameter(init_values * torch.ones(dim))
+
+    def forward(self, x: Tensor) -> Tensor:
+        return x.mul_(self.gamma) if self.inplace else x * self.gamma

depth_anything_v2/dinov2_layers/mlp.py

@@ -0,0 +1,41 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+# References:
+#   https://github.com/facebookresearch/dino/blob/master/vision_transformer.py
+#   https://github.com/rwightman/pytorch-image-models/tree/master/timm/layers/mlp.py
+
+
+from typing import Callable, Optional
+
+from torch import Tensor, nn
+
+
+class Mlp(nn.Module):
+    def __init__(
+        self,
+        in_features: int,
+        hidden_features: Optional[int] = None,
+        out_features: Optional[int] = None,
+        act_layer: Callable[..., nn.Module] = nn.GELU,
+        drop: float = 0.0,
+        bias: bool = True,
+    ) -> None:
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        self.fc1 = nn.Linear(in_features, hidden_features, bias=bias)
+        self.act = act_layer()
+        self.fc2 = nn.Linear(hidden_features, out_features, bias=bias)
+        self.drop = nn.Dropout(drop)
+
+    def forward(self, x: Tensor) -> Tensor:
+        x = self.fc1(x)
+        x = self.act(x)
+        x = self.drop(x)
+        x = self.fc2(x)
+        x = self.drop(x)
+        return x

depth_anything_v2/dinov2_layers/patch_embed.py

@@ -0,0 +1,89 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+# References:
+#   https://github.com/facebookresearch/dino/blob/master/vision_transformer.py
+#   https://github.com/rwightman/pytorch-image-models/tree/master/timm/layers/patch_embed.py
+
+from typing import Callable, Optional, Tuple, Union
+
+from torch import Tensor
+import torch.nn as nn
+
+
+def make_2tuple(x):
+    if isinstance(x, tuple):
+        assert len(x) == 2
+        return x
+
+    assert isinstance(x, int)
+    return (x, x)
+
+
+class PatchEmbed(nn.Module):
+    """
+    2D image to patch embedding: (B,C,H,W) -> (B,N,D)
+
+    Args:
+        img_size: Image size.
+        patch_size: Patch token size.
+        in_chans: Number of input image channels.
+        embed_dim: Number of linear projection output channels.
+        norm_layer: Normalization layer.
+    """
+
+    def __init__(
+        self,
+        img_size: Union[int, Tuple[int, int]] = 224,
+        patch_size: Union[int, Tuple[int, int]] = 16,
+        in_chans: int = 3,
+        embed_dim: int = 768,
+        norm_layer: Optional[Callable] = None,
+        flatten_embedding: bool = True,
+    ) -> None:
+        super().__init__()
+
+        image_HW = make_2tuple(img_size)
+        patch_HW = make_2tuple(patch_size)
+        patch_grid_size = (
+            image_HW[0] // patch_HW[0],
+            image_HW[1] // patch_HW[1],
+        )
+
+        self.img_size = image_HW
+        self.patch_size = patch_HW
+        self.patches_resolution = patch_grid_size
+        self.num_patches = patch_grid_size[0] * patch_grid_size[1]
+
+        self.in_chans = in_chans
+        self.embed_dim = embed_dim
+
+        self.flatten_embedding = flatten_embedding
+
+        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_HW, stride=patch_HW)
+        self.norm = norm_layer(embed_dim) if norm_layer else nn.Identity()
+
+    def forward(self, x: Tensor) -> Tensor:
+        _, _, H, W = x.shape
+        patch_H, patch_W = self.patch_size
+
+        assert H % patch_H == 0, f"Input image height {H} is not a multiple of patch height {patch_H}"
+        assert W % patch_W == 0, f"Input image width {W} is not a multiple of patch width: {patch_W}"
+
+        x = self.proj(x)  # B C H W
+        H, W = x.size(2), x.size(3)
+        x = x.flatten(2).transpose(1, 2)  # B HW C
+        x = self.norm(x)
+        if not self.flatten_embedding:
+            x = x.reshape(-1, H, W, self.embed_dim)  # B H W C
+        return x
+
+    def flops(self) -> float:
+        Ho, Wo = self.patches_resolution
+        flops = Ho * Wo * self.embed_dim * self.in_chans * (self.patch_size[0] * self.patch_size[1])
+        if self.norm is not None:
+            flops += Ho * Wo * self.embed_dim
+        return flops

depth_anything_v2/dinov2_layers/swiglu_ffn.py

@@ -0,0 +1,63 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from typing import Callable, Optional
+
+from torch import Tensor, nn
+import torch.nn.functional as F
+
+
+class SwiGLUFFN(nn.Module):
+    def __init__(
+        self,
+        in_features: int,
+        hidden_features: Optional[int] = None,
+        out_features: Optional[int] = None,
+        act_layer: Callable[..., nn.Module] = None,
+        drop: float = 0.0,
+        bias: bool = True,
+    ) -> None:
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        self.w12 = nn.Linear(in_features, 2 * hidden_features, bias=bias)
+        self.w3 = nn.Linear(hidden_features, out_features, bias=bias)
+
+    def forward(self, x: Tensor) -> Tensor:
+        x12 = self.w12(x)
+        x1, x2 = x12.chunk(2, dim=-1)
+        hidden = F.silu(x1) * x2
+        return self.w3(hidden)
+
+
+try:
+    from xformers.ops import SwiGLU
+
+    XFORMERS_AVAILABLE = True
+except ImportError:
+    SwiGLU = SwiGLUFFN
+    XFORMERS_AVAILABLE = False
+
+
+class SwiGLUFFNFused(SwiGLU):
+    def __init__(
+        self,
+        in_features: int,
+        hidden_features: Optional[int] = None,
+        out_features: Optional[int] = None,
+        act_layer: Callable[..., nn.Module] = None,
+        drop: float = 0.0,
+        bias: bool = True,
+    ) -> None:
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        hidden_features = (int(hidden_features * 2 / 3) + 7) // 8 * 8
+        super().__init__(
+            in_features=in_features,
+            hidden_features=hidden_features,
+            out_features=out_features,
+            bias=bias,
+        )

depth_anything_v2/dpt.py

@@ -0,0 +1,222 @@
+import cv2
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from torchvision.transforms import Compose
+
+from .dinov2 import DINOv2
+from .util.blocks import FeatureFusionBlock, _make_scratch
+from .util.transform import Resize, NormalizeImage, PrepareForNet
+
+
+def _make_fusion_block(features, use_bn, size=None):
+    return FeatureFusionBlock(
+        features,
+        nn.ReLU(False),
+        deconv=False,
+        bn=use_bn,
+        expand=False,
+        align_corners=True,
+        size=size,
+    )
+
+
+class ConvBlock(nn.Module):
+    def __init__(self, in_feature, out_feature):
+        super().__init__()
+        
+        self.conv_block = nn.Sequential(
+            nn.Conv2d(in_feature, out_feature, kernel_size=3, stride=1, padding=1),
+            nn.BatchNorm2d(out_feature),
+            nn.ReLU(True)
+        )
+    
+    def forward(self, x):
+        return self.conv_block(x)
+
+
+class DPTHead(nn.Module):
+    def __init__(
+        self, 
+        in_channels, 
+        features=256, 
+        use_bn=False, 
+        out_channels=[256, 512, 1024, 1024], 
+        use_clstoken=False
+    ):
+        super(DPTHead, self).__init__()
+        
+        self.use_clstoken = use_clstoken
+        
+        self.projects = nn.ModuleList([
+            nn.Conv2d(
+                in_channels=in_channels,
+                out_channels=out_channel,
+                kernel_size=1,
+                stride=1,
+                padding=0,
+            ) for out_channel in out_channels
+        ])
+        
+        self.resize_layers = nn.ModuleList([
+            nn.ConvTranspose2d(
+                in_channels=out_channels[0],
+                out_channels=out_channels[0],
+                kernel_size=4,
+                stride=4,
+                padding=0),
+            nn.ConvTranspose2d(
+                in_channels=out_channels[1],
+                out_channels=out_channels[1],
+                kernel_size=2,
+                stride=2,
+                padding=0),
+            nn.Identity(),
+            nn.Conv2d(
+                in_channels=out_channels[3],
+                out_channels=out_channels[3],
+                kernel_size=3,
+                stride=2,
+                padding=1)
+        ])
+        
+        if use_clstoken:
+            self.readout_projects = nn.ModuleList()
+            for _ in range(len(self.projects)):
+                self.readout_projects.append(
+                    nn.Sequential(
+                        nn.Linear(2 * in_channels, in_channels),
+                        nn.GELU()))
+        
+        self.scratch = _make_scratch(
+            out_channels,
+            features,
+            groups=1,
+            expand=False,
+        )
+        
+        self.scratch.stem_transpose = None
+        
+        self.scratch.refinenet1 = _make_fusion_block(features, use_bn)
+        self.scratch.refinenet2 = _make_fusion_block(features, use_bn)
+        self.scratch.refinenet3 = _make_fusion_block(features, use_bn)
+        self.scratch.refinenet4 = _make_fusion_block(features, use_bn)
+        
+        head_features_1 = features
+        head_features_2 = 32
+        
+        self.scratch.output_conv1 = nn.Conv2d(head_features_1, head_features_1 // 2, kernel_size=3, stride=1, padding=1)
+        self.scratch.output_conv2 = nn.Sequential(
+            nn.Conv2d(head_features_1 // 2, head_features_2, kernel_size=3, stride=1, padding=1),
+            nn.ReLU(True),
+            nn.Conv2d(head_features_2, 1, kernel_size=1, stride=1, padding=0),
+            nn.Sigmoid()
+        )
+    
+    def forward(self, out_features, patch_h, patch_w):
+        out = []
+        for i, x in enumerate(out_features):
+            if self.use_clstoken:
+                x, cls_token = x[0], x[1]
+                readout = cls_token.unsqueeze(1).expand_as(x)
+                x = self.readout_projects[i](torch.cat((x, readout), -1))
+            else:
+                x = x[0]
+            
+            x = x.permute(0, 2, 1).reshape((x.shape[0], x.shape[-1], patch_h, patch_w))
+            
+            x = self.projects[i](x)
+            x = self.resize_layers[i](x)
+            
+            out.append(x)
+        
+        layer_1, layer_2, layer_3, layer_4 = out
+        
+        layer_1_rn = self.scratch.layer1_rn(layer_1)
+        layer_2_rn = self.scratch.layer2_rn(layer_2)
+        layer_3_rn = self.scratch.layer3_rn(layer_3)
+        layer_4_rn = self.scratch.layer4_rn(layer_4)
+        
+        path_4 = self.scratch.refinenet4(layer_4_rn, size=layer_3_rn.shape[2:])        
+        path_3 = self.scratch.refinenet3(path_4, layer_3_rn, size=layer_2_rn.shape[2:])
+        path_2 = self.scratch.refinenet2(path_3, layer_2_rn, size=layer_1_rn.shape[2:])
+        path_1 = self.scratch.refinenet1(path_2, layer_1_rn)
+        
+        out = self.scratch.output_conv1(path_1)
+        out = F.interpolate(out, (int(patch_h * 14), int(patch_w * 14)), mode="bilinear", align_corners=True)
+        out = self.scratch.output_conv2(out)
+        
+        return out
+
+
+class DepthAnythingV2(nn.Module):
+    def __init__(
+        self, 
+        encoder='vitl', 
+        features=256, 
+        out_channels=[256, 512, 1024, 1024], 
+        use_bn=False, 
+        use_clstoken=False,
+        max_depth=20.0
+    ):
+        super(DepthAnythingV2, self).__init__()
+        
+        self.intermediate_layer_idx = {
+            'vits': [2, 5, 8, 11],
+            'vitb': [2, 5, 8, 11], 
+            'vitl': [4, 11, 17, 23], 
+            'vitg': [9, 19, 29, 39]
+        }
+        
+        self.max_depth = max_depth
+        
+        self.encoder = encoder
+        self.pretrained = DINOv2(model_name=encoder)
+        
+        self.depth_head = DPTHead(self.pretrained.embed_dim, features, use_bn, out_channels=out_channels, use_clstoken=use_clstoken)
+    
+    def forward(self, x):
+        patch_h, patch_w = x.shape[-2] // 14, x.shape[-1] // 14
+        
+        features = self.pretrained.get_intermediate_layers(x, self.intermediate_layer_idx[self.encoder], return_class_token=True)
+        
+        depth = self.depth_head(features, patch_h, patch_w) * self.max_depth
+        
+        return depth.squeeze(1)
+    
+    @torch.no_grad()
+    def infer_image(self, raw_image, input_size=518):
+        image, (h, w) = self.image2tensor(raw_image, input_size)
+        
+        depth = self.forward(image)
+        
+        depth = F.interpolate(depth[:, None], (h, w), mode="bilinear", align_corners=True)[0, 0]
+        
+        return depth.cpu().numpy()
+    
+    def image2tensor(self, raw_image, input_size=518):        
+        transform = Compose([
+            Resize(
+                width=input_size,
+                height=input_size,
+                resize_target=False,
+                keep_aspect_ratio=True,
+                ensure_multiple_of=14,
+                resize_method='lower_bound',
+                image_interpolation_method=cv2.INTER_CUBIC,
+            ),
+            NormalizeImage(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
+            PrepareForNet(),
+        ])
+        
+        h, w = raw_image.shape[:2]
+        
+        image = cv2.cvtColor(raw_image, cv2.COLOR_BGR2RGB) / 255.0
+        
+        image = transform({'image': image})['image']
+        image = torch.from_numpy(image).unsqueeze(0)
+        
+        DEVICE = 'cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu'
+        image = image.to(DEVICE)
+        
+        return image, (h, w)

depth_anything_v2/util/blocks.py

@@ -0,0 +1,148 @@
+import torch.nn as nn
+
+
+def _make_scratch(in_shape, out_shape, groups=1, expand=False):
+    scratch = nn.Module()
+
+    out_shape1 = out_shape
+    out_shape2 = out_shape
+    out_shape3 = out_shape
+    if len(in_shape) >= 4:
+        out_shape4 = out_shape
+
+    if expand:
+        out_shape1 = out_shape
+        out_shape2 = out_shape * 2
+        out_shape3 = out_shape * 4
+        if len(in_shape) >= 4:
+            out_shape4 = out_shape * 8
+
+    scratch.layer1_rn = nn.Conv2d(in_shape[0], out_shape1, kernel_size=3, stride=1, padding=1, bias=False, groups=groups)
+    scratch.layer2_rn = nn.Conv2d(in_shape[1], out_shape2, kernel_size=3, stride=1, padding=1, bias=False, groups=groups)
+    scratch.layer3_rn = nn.Conv2d(in_shape[2], out_shape3, kernel_size=3, stride=1, padding=1, bias=False, groups=groups)
+    if len(in_shape) >= 4:
+        scratch.layer4_rn = nn.Conv2d(in_shape[3], out_shape4, kernel_size=3, stride=1, padding=1, bias=False, groups=groups)
+
+    return scratch
+
+
+class ResidualConvUnit(nn.Module):
+    """Residual convolution module.
+    """
+
+    def __init__(self, features, activation, bn):
+        """Init.
+
+        Args:
+            features (int): number of features
+        """
+        super().__init__()
+
+        self.bn = bn
+
+        self.groups=1
+
+        self.conv1 = nn.Conv2d(features, features, kernel_size=3, stride=1, padding=1, bias=True, groups=self.groups)
+        
+        self.conv2 = nn.Conv2d(features, features, kernel_size=3, stride=1, padding=1, bias=True, groups=self.groups)
+
+        if self.bn == True:
+            self.bn1 = nn.BatchNorm2d(features)
+            self.bn2 = nn.BatchNorm2d(features)
+
+        self.activation = activation
+
+        self.skip_add = nn.quantized.FloatFunctional()
+
+    def forward(self, x):
+        """Forward pass.
+
+        Args:
+            x (tensor): input
+
+        Returns:
+            tensor: output
+        """
+        
+        out = self.activation(x)
+        out = self.conv1(out)
+        if self.bn == True:
+            out = self.bn1(out)
+       
+        out = self.activation(out)
+        out = self.conv2(out)
+        if self.bn == True:
+            out = self.bn2(out)
+
+        if self.groups > 1:
+            out = self.conv_merge(out)
+
+        return self.skip_add.add(out, x)
+
+
+class FeatureFusionBlock(nn.Module):
+    """Feature fusion block.
+    """
+
+    def __init__(
+        self, 
+        features, 
+        activation, 
+        deconv=False, 
+        bn=False, 
+        expand=False, 
+        align_corners=True,
+        size=None
+    ):
+        """Init.
+        
+        Args:
+            features (int): number of features
+        """
+        super(FeatureFusionBlock, self).__init__()
+
+        self.deconv = deconv
+        self.align_corners = align_corners
+
+        self.groups=1
+
+        self.expand = expand
+        out_features = features
+        if self.expand == True:
+            out_features = features // 2
+        
+        self.out_conv = nn.Conv2d(features, out_features, kernel_size=1, stride=1, padding=0, bias=True, groups=1)
+
+        self.resConfUnit1 = ResidualConvUnit(features, activation, bn)
+        self.resConfUnit2 = ResidualConvUnit(features, activation, bn)
+        
+        self.skip_add = nn.quantized.FloatFunctional()
+
+        self.size=size
+
+    def forward(self, *xs, size=None):
+        """Forward pass.
+
+        Returns:
+            tensor: output
+        """
+        output = xs[0]
+
+        if len(xs) == 2:
+            res = self.resConfUnit1(xs[1])
+            output = self.skip_add.add(output, res)
+
+        output = self.resConfUnit2(output)
+
+        if (size is None) and (self.size is None):
+            modifier = {"scale_factor": 2}
+        elif size is None:
+            modifier = {"size": self.size}
+        else:
+            modifier = {"size": size}
+
+        output = nn.functional.interpolate(output, **modifier, mode="bilinear", align_corners=self.align_corners)
+        
+        output = self.out_conv(output)
+
+        return output

depth_anything_v2/util/transform.py

@@ -0,0 +1,158 @@
+import numpy as np
+import cv2
+
+
+class Resize(object):
+    """Resize sample to given size (width, height).
+    """
+
+    def __init__(
+        self,
+        width,
+        height,
+        resize_target=True,
+        keep_aspect_ratio=False,
+        ensure_multiple_of=1,
+        resize_method="lower_bound",
+        image_interpolation_method=cv2.INTER_AREA,
+    ):
+        """Init.
+
+        Args:
+            width (int): desired output width
+            height (int): desired output height
+            resize_target (bool, optional):
+                True: Resize the full sample (image, mask, target).
+                False: Resize image only.
+                Defaults to True.
+            keep_aspect_ratio (bool, optional):
+                True: Keep the aspect ratio of the input sample.
+                Output sample might not have the given width and height, and
+                resize behaviour depends on the parameter 'resize_method'.
+                Defaults to False.
+            ensure_multiple_of (int, optional):
+                Output width and height is constrained to be multiple of this parameter.
+                Defaults to 1.
+            resize_method (str, optional):
+                "lower_bound": Output will be at least as large as the given size.
+                "upper_bound": Output will be at max as large as the given size. (Output size might be smaller than given size.)
+                "minimal": Scale as least as possible.  (Output size might be smaller than given size.)
+                Defaults to "lower_bound".
+        """
+        self.__width = width
+        self.__height = height
+
+        self.__resize_target = resize_target
+        self.__keep_aspect_ratio = keep_aspect_ratio
+        self.__multiple_of = ensure_multiple_of
+        self.__resize_method = resize_method
+        self.__image_interpolation_method = image_interpolation_method
+
+    def constrain_to_multiple_of(self, x, min_val=0, max_val=None):
+        y = (np.round(x / self.__multiple_of) * self.__multiple_of).astype(int)
+
+        if max_val is not None and y > max_val:
+            y = (np.floor(x / self.__multiple_of) * self.__multiple_of).astype(int)
+
+        if y < min_val:
+            y = (np.ceil(x / self.__multiple_of) * self.__multiple_of).astype(int)
+
+        return y
+
+    def get_size(self, width, height):
+        # determine new height and width
+        scale_height = self.__height / height
+        scale_width = self.__width / width
+
+        if self.__keep_aspect_ratio:
+            if self.__resize_method == "lower_bound":
+                # scale such that output size is lower bound
+                if scale_width > scale_height:
+                    # fit width
+                    scale_height = scale_width
+                else:
+                    # fit height
+                    scale_width = scale_height
+            elif self.__resize_method == "upper_bound":
+                # scale such that output size is upper bound
+                if scale_width < scale_height:
+                    # fit width
+                    scale_height = scale_width
+                else:
+                    # fit height
+                    scale_width = scale_height
+            elif self.__resize_method == "minimal":
+                # scale as least as possbile
+                if abs(1 - scale_width) < abs(1 - scale_height):
+                    # fit width
+                    scale_height = scale_width
+                else:
+                    # fit height
+                    scale_width = scale_height
+            else:
+                raise ValueError(f"resize_method {self.__resize_method} not implemented")
+
+        if self.__resize_method == "lower_bound":
+            new_height = self.constrain_to_multiple_of(scale_height * height, min_val=self.__height)
+            new_width = self.constrain_to_multiple_of(scale_width * width, min_val=self.__width)
+        elif self.__resize_method == "upper_bound":
+            new_height = self.constrain_to_multiple_of(scale_height * height, max_val=self.__height)
+            new_width = self.constrain_to_multiple_of(scale_width * width, max_val=self.__width)
+        elif self.__resize_method == "minimal":
+            new_height = self.constrain_to_multiple_of(scale_height * height)
+            new_width = self.constrain_to_multiple_of(scale_width * width)
+        else:
+            raise ValueError(f"resize_method {self.__resize_method} not implemented")
+
+        return (new_width, new_height)
+
+    def __call__(self, sample):
+        width, height = self.get_size(sample["image"].shape[1], sample["image"].shape[0])
+        
+        # resize sample
+        sample["image"] = cv2.resize(sample["image"], (width, height), interpolation=self.__image_interpolation_method)
+
+        if self.__resize_target:
+            if "depth" in sample:
+                sample["depth"] = cv2.resize(sample["depth"], (width, height), interpolation=cv2.INTER_NEAREST)
+                
+            if "mask" in sample:
+                sample["mask"] = cv2.resize(sample["mask"].astype(np.float32), (width, height), interpolation=cv2.INTER_NEAREST)
+        
+        return sample
+
+
+class NormalizeImage(object):
+    """Normlize image by given mean and std.
+    """
+
+    def __init__(self, mean, std):
+        self.__mean = mean
+        self.__std = std
+
+    def __call__(self, sample):
+        sample["image"] = (sample["image"] - self.__mean) / self.__std
+
+        return sample
+
+
+class PrepareForNet(object):
+    """Prepare sample for usage as network input.
+    """
+
+    def __init__(self):
+        pass
+
+    def __call__(self, sample):
+        image = np.transpose(sample["image"], (2, 0, 1))
+        sample["image"] = np.ascontiguousarray(image).astype(np.float32)
+
+        if "depth" in sample:
+            depth = sample["depth"].astype(np.float32)
+            sample["depth"] = np.ascontiguousarray(depth)
+        
+        if "mask" in sample:
+            sample["mask"] = sample["mask"].astype(np.float32)
+            sample["mask"] = np.ascontiguousarray(sample["mask"])
+        
+        return sample

dist_train.sh

@@ -0,0 +1,26 @@
+#!/bin/bash
+now=$(date +"%Y%m%d_%H%M%S")
+
+epoch=120
+bs=1
+gpus=1
+lr=0.000005
+encoder=vitl
+dataset=pbr  # Changed default to pbr
+img_size=512
+min_depth=0.0
+max_depth=1.0  # Changed to 1.0 for normalized depth
+pretrained_from=../checkpoints/depth_anything_v2_${encoder}.pth
+save_path=exp/pbr  # Changed to pbr
+
+mkdir -p $save_path
+
+python3 -m torch.distributed.launch \
+    --nproc_per_node=$gpus \
+    --nnodes 1 \
+    --node_rank=0 \
+    --master_addr=localhost \
+    --master_port=20596 \
+    train.py --epoch $epoch --encoder $encoder --bs $bs --lr $lr --save-path $save_path --dataset $dataset \
+    --img-size $img_size --min-depth $min_depth --max-depth $max_depth --pretrained-from $pretrained_from \
+    --port 20596 2>&1 | tee -a $save_path/$now.log

requirements.txt

@@ -0,0 +1,5 @@
+matplotlib
+opencv-python
+open3d
+torch
+torchvision

run.py

@@ -0,0 +1,81 @@
+import argparse
+import cv2
+import glob
+import matplotlib
+import numpy as np
+import os
+import torch
+
+from depth_anything_v2.dpt import DepthAnythingV2
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description='Depth Anything V2 Metric Depth Estimation')
+    
+    parser.add_argument('--img-path', type=str)
+    parser.add_argument('--input-size', type=int, default=518)
+    parser.add_argument('--outdir', type=str, default='./vis_depth')
+    
+    parser.add_argument('--encoder', type=str, default='vitl', choices=['vits', 'vitb', 'vitl', 'vitg'])
+    parser.add_argument('--load-from', type=str, default='checkpoints/depth_anything_v2_metric_hypersim_vitl.pth')
+    parser.add_argument('--max-depth', type=float, default=20)
+    
+    parser.add_argument('--save-numpy', dest='save_numpy', action='store_true', help='save the model raw output')
+    parser.add_argument('--pred-only', dest='pred_only', action='store_true', help='only display the prediction')
+    parser.add_argument('--grayscale', dest='grayscale', action='store_true', help='do not apply colorful palette')
+    
+    args = parser.parse_args()
+    
+    DEVICE = 'cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu'
+    
+    model_configs = {
+        'vits': {'encoder': 'vits', 'features': 64, 'out_channels': [48, 96, 192, 384]},
+        'vitb': {'encoder': 'vitb', 'features': 128, 'out_channels': [96, 192, 384, 768]},
+        'vitl': {'encoder': 'vitl', 'features': 256, 'out_channels': [256, 512, 1024, 1024]},
+        'vitg': {'encoder': 'vitg', 'features': 384, 'out_channels': [1536, 1536, 1536, 1536]}
+    }
+    
+    depth_anything = DepthAnythingV2(**{**model_configs[args.encoder], 'max_depth': args.max_depth})
+    depth_anything.load_state_dict(torch.load(args.load_from, map_location='cpu'))
+    depth_anything = depth_anything.to(DEVICE).eval()
+    
+    if os.path.isfile(args.img_path):
+        if args.img_path.endswith('txt'):
+            with open(args.img_path, 'r') as f:
+                filenames = f.read().splitlines()
+        else:
+            filenames = [args.img_path]
+    else:
+        filenames = glob.glob(os.path.join(args.img_path, '**/*'), recursive=True)
+    
+    os.makedirs(args.outdir, exist_ok=True)
+    
+    cmap = matplotlib.colormaps.get_cmap('Spectral')
+    
+    for k, filename in enumerate(filenames):
+        print(f'Progress {k+1}/{len(filenames)}: {filename}')
+        
+        raw_image = cv2.imread(filename)
+        
+        depth = depth_anything.infer_image(raw_image, args.input_size)
+        
+        if args.save_numpy:
+            output_path = os.path.join(args.outdir, os.path.splitext(os.path.basename(filename))[0] + '_raw_depth_meter.npy')
+            np.save(output_path, depth)
+        
+        depth = (depth - depth.min()) / (depth.max() - depth.min()) * 255.0
+        depth = depth.astype(np.uint8)
+        
+        if args.grayscale:
+            depth = np.repeat(depth[..., np.newaxis], 3, axis=-1)
+        else:
+            depth = (cmap(depth)[:, :, :3] * 255)[:, :, ::-1].astype(np.uint8)
+        
+        output_path = os.path.join(args.outdir, os.path.splitext(os.path.basename(filename))[0] + '.png')
+        if args.pred_only:
+            cv2.imwrite(output_path, depth)
+        else:
+            split_region = np.ones((raw_image.shape[0], 50, 3), dtype=np.uint8) * 255
+            combined_result = cv2.hconcat([raw_image, split_region, depth])
+            
+            cv2.imwrite(output_path, combined_result)

run2.py

@@ -0,0 +1,54 @@
+import cv2
+import torch
+import numpy as np
+
+from depth_anything_v2.dpt import DepthAnythingV2
+
+model_configs = {
+    'vits': {'encoder': 'vits', 'features': 64, 'out_channels': [48, 96, 192, 384]},
+    'vitb': {'encoder': 'vitb', 'features': 128, 'out_channels': [96, 192, 384, 768]},
+    'vitl': {'encoder': 'vitl', 'features': 256, 'out_channels': [256, 512, 1024, 1024]}
+}
+
+encoder = 'vitl' # or 'vits', 'vitb'
+dataset = 'pbr' # 'hypersim' for indoor model, 'vkitti' for outdoor model
+max_depth = 1 # 20 for indoor model, 80 for outdoor model
+
+model = DepthAnythingV2(**{**model_configs[encoder], 'max_depth': max_depth})
+
+# Load checkpoint and handle unexpected keys
+checkpoint = torch.load(f'checkpoints/model2.pth', map_location='cpu')
+print("Keys in checkpoint:", checkpoint.keys())
+
+# Skip unexpected keys
+expected_keys = ['model']
+state_dict = {}
+for key in checkpoint.keys():
+    if key not in ['optimizer', 'epoch', 'previous_best']:
+        state_dict = checkpoint[key]
+        print(f"Using weights from key: {key}")
+    else:
+        print(f"Skipping unexpected key: {key}")
+
+# Handle module prefix if present
+my_state_dict = {}
+for key in state_dict.keys():
+    new_key = key.replace('module.', '')
+    my_state_dict[new_key] = state_dict[key]
+
+model.load_state_dict(my_state_dict)
+model.eval()
+
+raw_img = cv2.imread('image.jpg')
+depth = model.infer_image(raw_img) # HxW depth map in meters in numpy
+
+# Normalize depth for visualization (0-255)
+depth_normalized = ((depth - depth.min()) / (depth.max() - depth.min()) * 255).astype(np.uint8)
+
+# Apply colormap for better visualization
+depth_colormap = cv2.applyColorMap(depth_normalized, cv2.COLORMAP_INFERNO)
+
+# Save both raw depth and colored depth
+cv2.imwrite('depth_raw.png', depth_normalized)
+cv2.imwrite('depth_colored.png', depth_colormap)
+print("Saved depth maps as 'depth_raw.png' and 'depth_colored.png'")

train.py

@@ -0,0 +1,253 @@
+import argparse
+import logging
+import os
+import pprint
+import random
+
+import numpy as np
+import torch
+import torch.backends.cudnn as cudnn
+import torch.distributed as dist
+from torch.utils.data import DataLoader
+from torch.optim import AdamW
+import torch.nn.functional as F
+from torch.utils.tensorboard import SummaryWriter
+
+from dataset.hypersim import Hypersim
+from dataset.kitti import KITTI
+from dataset.vkitti2 import VKITTI2
+from dataset.pbr import PBRDataset
+from depth_anything_v2.dpt import DepthAnythingV2
+from util.dist_helper import setup_distributed
+from util.loss import SiLogLoss
+from util.metric import eval_depth
+from util.utils import init_log
+
+
+def rotate_sample(img, depth, valid_mask, angle):
+    """
+    Rotate image, depth map and valid mask by specified angle (90, 180, or 270 degrees)
+    """
+    k = angle // 90  # k=1 for 90°, k=2 for 180°, k=3 for 270°
+    img = torch.rot90(img, k, dims=[-2, -1])
+    depth = torch.rot90(depth, k, dims=[-2, -1])
+    valid_mask = torch.rot90(valid_mask, k, dims=[-2, -1])
+    return img, depth, valid_mask
+
+
+parser = argparse.ArgumentParser(description='Depth Anything V2 for Metric Depth Estimation')
+
+parser.add_argument('--encoder', default='vitl', choices=['vits', 'vitb', 'vitl', 'vitg'])
+parser.add_argument('--dataset', default='hypersim', choices=['hypersim', 'vkitti', 'pbr'])
+parser.add_argument('--img-size', default=512, type=int)
+parser.add_argument('--min-depth', default=0.001, type=float)
+parser.add_argument('--max-depth', default=1.0, type=float)
+parser.add_argument('--epochs', default=40, type=int)
+parser.add_argument('--bs', default=2, type=int)
+parser.add_argument('--lr', default=0.000005, type=float)
+parser.add_argument('--pretrained-from', type=str)
+parser.add_argument('--save-path', type=str, required=True)
+parser.add_argument('--local-rank', default=0, type=int)
+parser.add_argument('--port', default=None, type=int)
+parser.add_argument('--flip-prob', default=0.5, type=float, help='Probability of horizontal flip')
+parser.add_argument('--rotate-prob', default=0.75, type=float, help='Probability of rotation')
+
+
+def main():
+    args = parser.parse_args()
+    rank = 0
+    
+    logger = init_log('global', logging.INFO)
+    logger.propagate = 0
+    
+    rank, world_size = setup_distributed(port=args.port)
+    
+    if rank == 0:
+        all_args = {**vars(args), 'ngpus': world_size}
+        logger.info('{}\n'.format(pprint.pformat(all_args)))
+        writer = SummaryWriter(args.save_path)
+    
+    cudnn.enabled = True
+    cudnn.benchmark = True
+    
+    size = (args.img_size, args.img_size)
+    if args.dataset == 'hypersim':
+        trainset = Hypersim('dataset/splits/hypersim/train.txt', 'train', size=size)
+    elif args.dataset == 'vkitti':
+        trainset = VKITTI2('dataset/splits/vkitti2/train.txt', 'train', size=size)
+    elif args.dataset == 'pbr':
+        trainset = PBRDataset('dataset/splits/pbr/train.txt', 'train', size=size)
+    else:
+        raise NotImplementedError
+    trainsampler = torch.utils.data.distributed.DistributedSampler(trainset)
+    trainloader = DataLoader(trainset, batch_size=args.bs, pin_memory=True, num_workers=4, drop_last=True, sampler=trainsampler)
+    
+    if args.dataset == 'hypersim':
+        valset = Hypersim('dataset/splits/hypersim/val.txt', 'val', size=size)
+    elif args.dataset == 'vkitti':
+        valset = KITTI('dataset/splits/kitti/val.txt', 'val', size=size)
+    elif args.dataset == 'pbr':
+        valset = PBRDataset('dataset/splits/pbr/val.txt', 'val', size=size)
+    else:
+        raise NotImplementedError
+    valsampler = torch.utils.data.distributed.DistributedSampler(valset)
+    valloader = DataLoader(valset, batch_size=1, pin_memory=True, num_workers=4, drop_last=True, sampler=valsampler)
+    
+    local_rank = int(os.environ["LOCAL_RANK"])
+    
+    model_configs = {
+        'vits': {'encoder': 'vits', 'features': 64, 'out_channels': [48, 96, 192, 384]},
+        'vitb': {'encoder': 'vitb', 'features': 128, 'out_channels': [96, 192, 384, 768]},
+        'vitl': {'encoder': 'vitl', 'features': 256, 'out_channels': [256, 512, 1024, 1024]},
+        'vitg': {'encoder': 'vitg', 'features': 384, 'out_channels': [1536, 1536, 1536, 1536]}
+    }
+    model = DepthAnythingV2(**{**model_configs[args.encoder], 'max_depth': args.max_depth})
+    
+    if args.pretrained_from:
+        model.load_state_dict({k: v for k, v in torch.load(args.pretrained_from, map_location='cpu').items() if 'pretrained' in k}, strict=False)
+    
+    model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(model)
+    model.cuda(local_rank)
+    model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[local_rank], broadcast_buffers=False,
+                                                     output_device=local_rank, find_unused_parameters=True)
+    
+    criterion = SiLogLoss().cuda(local_rank)
+    
+    optimizer = AdamW([{'params': [param for name, param in model.named_parameters() if 'pretrained' in name], 'lr': args.lr},
+                      {'params': [param for name, param in model.named_parameters() if 'pretrained' not in name], 'lr': args.lr * 10.0}],
+                     lr=args.lr, betas=(0.9, 0.999), weight_decay=0.01)
+    
+    total_iters = args.epochs * len(trainloader)
+    
+    # Initialize previous_best dictionary here
+    previous_best = {'d1': 0, 'd2': 0, 'd3': 0, 'abs_rel': 100, 'sq_rel': 100, 
+                    'rmse': 100, 'rmse_log': 100, 'log10': 100, 'silog': 100}
+    
+    # Load checkpoint if exists
+    if os.path.exists(os.path.join(args.save_path, 'latest.pth')):
+        checkpoint = torch.load(os.path.join(args.save_path, 'latest.pth'), map_location='cpu')
+        model.load_state_dict(checkpoint['model'])
+        optimizer.load_state_dict(checkpoint['optimizer'])
+        previous_best = checkpoint['previous_best']
+        start_epoch = checkpoint['epoch'] + 1
+    else:
+        start_epoch = 0
+
+    for epoch in range(start_epoch, args.epochs):
+        if rank == 0:
+            logger.info('===========> Epoch: {:}/{:}, d1: {:.3f}, d2: {:.3f}, d3: {:.3f}'.format(
+                epoch, args.epochs, previous_best['d1'], previous_best['d2'], previous_best['d3']))
+            logger.info('===========> Epoch: {:}/{:}, abs_rel: {:.3f}, sq_rel: {:.3f}, rmse: {:.3f}, rmse_log: {:.3f}, '
+                      'log10: {:.3f}, silog: {:.3f}'.format(
+                          epoch, args.epochs, previous_best['abs_rel'], previous_best['sq_rel'], previous_best['rmse'],
+                          previous_best['rmse_log'], previous_best['log10'], previous_best['silog']))
+        
+        trainloader.sampler.set_epoch(epoch + 1)
+        
+        model.train()
+        total_loss = 0
+        
+        for i, sample in enumerate(trainloader):
+            optimizer.zero_grad()
+            
+            img, depth, valid_mask = sample['image'].cuda(), sample['depth'].cuda(), sample['valid_mask'].cuda()
+            
+            # Apply random horizontal flip
+            if random.random() < 0.5:
+                img = img.flip(-1)
+                depth = depth.flip(-1)
+                valid_mask = valid_mask.flip(-1)
+            
+            # Apply random rotation augmentation
+            rotation_prob = random.random()
+            if rotation_prob < 0.75:
+                if rotation_prob < 0.25:  # 90°
+                    img, depth, valid_mask = rotate_sample(img, depth, valid_mask, 90)
+                elif rotation_prob < 0.5:  # 180°
+                    img, depth, valid_mask = rotate_sample(img, depth, valid_mask, 180)
+                else:  # 270°
+                    img, depth, valid_mask = rotate_sample(img, depth, valid_mask, 270)
+            
+            pred = model(img)
+            
+            loss = criterion(pred, depth, (valid_mask == 1) & (depth >= args.min_depth) & (depth <= args.max_depth))
+            
+            loss.backward()
+            optimizer.step()
+            
+            total_loss += loss.item()
+            
+            iters = epoch * len(trainloader) + i
+            
+            lr = args.lr * (1 - iters / total_iters) ** 0.9
+            
+            optimizer.param_groups[0]["lr"] = lr
+            optimizer.param_groups[1]["lr"] = lr * 10.0
+            
+            if rank == 0:
+                writer.add_scalar('train/loss', loss.item(), iters)
+            
+            if rank == 0 and i % 100 == 0:
+                logger.info('Iter: {}/{}, LR: {:.7f}, Loss: {:.3f}'.format(
+                    i, len(trainloader), optimizer.param_groups[0]['lr'], loss.item()))
+        
+        model.eval()
+        
+        results = {'d1': torch.tensor([0.0]).cuda(), 'd2': torch.tensor([0.0]).cuda(), 'd3': torch.tensor([0.0]).cuda(),
+                  'abs_rel': torch.tensor([0.0]).cuda(), 'sq_rel': torch.tensor([0.0]).cuda(), 'rmse': torch.tensor([0.0]).cuda(),
+                  'rmse_log': torch.tensor([0.0]).cuda(), 'log10': torch.tensor([0.0]).cuda(), 'silog': torch.tensor([0.0]).cuda()}
+        nsamples = torch.tensor([0.0]).cuda()
+        
+        for i, sample in enumerate(valloader):
+            img, depth, valid_mask = sample['image'].cuda().float(), sample['depth'].cuda()[0], sample['valid_mask'].cuda()[0]
+            
+            with torch.no_grad():
+                pred = model(img)
+                pred = F.interpolate(pred[:, None], depth.shape[-2:], mode='bilinear', align_corners=True)[0, 0]
+            
+            valid_mask = (valid_mask == 1) & (depth >= args.min_depth) & (depth <= args.max_depth)
+            
+            if valid_mask.sum() < 10:
+                continue
+            
+            cur_results = eval_depth(pred[valid_mask], depth[valid_mask])
+            
+            for k in results.keys():
+                results[k] += cur_results[k]
+            nsamples += 1
+        
+        torch.distributed.barrier()
+        
+        for k in results.keys():
+            dist.reduce(results[k], dst=0)
+        dist.reduce(nsamples, dst=0)
+        
+        if rank == 0:
+            logger.info('==========================================================================================')
+            logger.info('{:>8}, {:>8}, {:>8}, {:>8}, {:>8}, {:>8}, {:>8}, {:>8}, {:>8}'.format(*tuple(results.keys())))
+            logger.info('{:8.3f}, {:8.3f}, {:8.3f}, {:8.3f}, {:8.3f}, {:8.3f}, {:8.3f}, {:8.3f}, {:8.3f}'.format(
+                *tuple([(v / nsamples).item() for v in results.values()])))
+            logger.info('==========================================================================================')
+            print()
+            
+            for name, metric in results.items():
+                writer.add_scalar(f'eval/{name}', (metric / nsamples).item(), epoch)
+        
+        for k in results.keys():
+            if k in ['d1', 'd2', 'd3']:
+                previous_best[k] = max(previous_best[k], (results[k] / nsamples).item())
+            else:
+                previous_best[k] = min(previous_best[k], (results[k] / nsamples).item())
+        
+        if rank == 0:
+            checkpoint = {
+                'model': model.state_dict(),
+                'optimizer': optimizer.state_dict(),
+                'epoch': epoch,
+                'previous_best': previous_best,
+            }
+            torch.save(checkpoint, os.path.join(args.save_path, 'latest.pth'))
+
+
+if __name__ == '__main__':
+    main()