AEmotionStudio commited on
Commit
b6f61c4
·
verified ·
1 Parent(s): 89fcbc1

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +64 -0
README.md ADDED
@@ -0,0 +1,64 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: depth-anything/Video-Depth-Anything-Large
4
+ tags:
5
+ - video
6
+ - depth-estimation
7
+ - monocular-depth
8
+ - safetensors
9
+ ---
10
+
11
+ # Video Depth Anything — Large
12
+
13
+ Mirror of [depth-anything/Video-Depth-Anything-Large](https://huggingface.co/depth-anything/Video-Depth-Anything-Large) for use with [ComfyUI-FFMPEGA](https://github.com/AEmotionStudio/ComfyUI-FFMPEGA).
14
+
15
+ ## What is Video Depth Anything?
16
+
17
+ Video Depth Anything is a state-of-the-art model for **temporally consistent monocular depth estimation** in videos. It extends [Depth Anything V2](https://github.com/DepthAnything/Depth-Anything-V2) with temporal modules for smooth, flicker-free depth maps across video frames.
18
+
19
+ Key features:
20
+ - **Temporal consistency** — smooth depth maps without frame-to-frame flickering
21
+ - **Multiple encoder sizes** — Small (335.3M), Base, and Large variants
22
+ - **Apache 2.0 license** — fully open source
23
+ - **Colormap output** — supports multiple colormap visualizations (inferno, magma, plasma, etc.)
24
+
25
+ ## Files
26
+
27
+ ```
28
+ model.safetensors
29
+ config.json
30
+ ```
31
+
32
+ ## Usage
33
+
34
+ ### With ComfyUI-FFMPEGA (recommended)
35
+
36
+ 1. Set `no_llm_mode` to `video_depth` on the FFMPEG Agent node
37
+ 2. Select encoder size (`small`, `base`, `large`) under Advanced Options
38
+ 3. Choose colormap for visualization
39
+ 4. The model auto-downloads on first use
40
+
41
+ ### Manual download
42
+
43
+ ```bash
44
+ huggingface-cli download AEmotionStudio/Video-Depth-Anything-Large --local-dir ./video_depth_anything
45
+ ```
46
+
47
+ ## Available Sizes
48
+
49
+ | Variant | Parameters | Size | Speed |
50
+ | :--- | :--- | :--- | :--- |
51
+ | **Small** | 24.8M | ~102 MB | Fastest |
52
+ | **Base** | 97.5M | ~390 MB | Balanced |
53
+ | **Large** | 335.3M | ~670 MB | Best quality |
54
+
55
+ ## License
56
+
57
+ Apache 2.0 — see the upstream repository for full license terms.
58
+
59
+ ## Credits
60
+
61
+ - Original model by: [Depth Anything team](https://github.com/DepthAnything/Video-Depth-Anything)
62
+ - Paper: *"Video Depth Anything: Consistent Depth Estimation for Super-Long Videos"*
63
+ - Upstream HuggingFace: [depth-anything/Video-Depth-Anything-Large](https://huggingface.co/depth-anything/Video-Depth-Anything-Large)
64
+ - Redistributed by: [Æmotion Studio](https://huggingface.co/AEmotionStudio) for use with [ComfyUI-FFMPEGA](https://github.com/AEmotionStudio/ComfyUI-FFMPEGA)