Video Depth Anything (fastforks)
Video Depth Anything (Chen et al., CVPR 2025): temporally consistent depth of long videos from the Depth Anything V2 encoder and a DPT head with temporal attention, run in overlapping 32-frame windows; relative (Small) and metric (Metric Small) checkpoints.
This repo is the model card only: it holds no weights. fastforks downloads the authors' weights from the links below and checks them against the SHA-256 in the model card. Code: https://github.com/kleinicke/fastforks. Kernels for the kernels backend: https://huggingface.co/fastforks/kernels. Original code: https://github.com/DepthAnything/Video-Depth-Anything at 4f5ae23172ba.
Author weight repos: https://huggingface.co/depth-anything/Metric-Video-Depth-Anything-Small, https://huggingface.co/depth-anything/Video-Depth-Anything-Small.
Task: depth. Tier: experimental. Outputs: inverse_depth, depth. Backends: portable, kernels.
Input: one RGB image. Output: a depth or inverse-depth map. kind (a card field, result.kind) says what the numbers mean: metric = metres; scale_invariant, affine_invariant, affine_invariant_log (log depth) and affine_invariant_inverse are relative (unknown scale or shift, not metres). Metric models that use the focal length take model(image, intrinsics=K) (K [3, 3], pixels of the input): metric3d-vit-large and da3metric-large need it for correct metres (da3metric-large: metres = output x focal / 300; without intrinsics= a 300 px focal length is assumed, wrong for a real camera). unidepth-v2-vitl14, unik3d-vitl, depth-pro and MoGe estimate the camera themselves (given intrinsics= replace fx); zoedepth-nk and the relative models (da3mono-large too) take none. Normals-only models (kind none) return result.normals [B, H, W, 3] (unit vectors, camera frame: x right, y down, z forward).
Variants
Weights are the authors' files (not hosted here). SHA-256 shortened to 12 hex digits; the full hashes are in card.toml.
| variant | weights (author link) | SHA-256 | weights licence | commercial use |
|---|---|---|---|---|
video-depth-anything-metric-s@1 |
metric_video_depth_anything_vits.pth | 3c28432b4e1f |
Apache-2.0 | yes |
video-depth-anything-s@1 |
video_depth_anything_vits.pth | 13379300b739 |
Apache-2.0 | yes |
Licence
Original-code strings, recorded as they are (research-use package; check them before any other use). Code: Apache-2.0. Licence files: LICENSE.
Training data: Depth Anything V2 Small encoder; the temporal head on synthetic video depth and pseudo-labelled real videos and images per the paper (arXiv 2501.12375).
- Licence note: Only the Small checkpoints (relative and metric) are Apache-2.0; the Base and Large checkpoints are CC-BY-NC-4.0 and not included.
- Licence note: The metric model is trained on Virtual KITTI 2 (CC BY-NC-SA 3.0) and IRS per the upstream README (a note only).
Licence evidence: https://github.com/DepthAnything/Video-Depth-Anything/blob/4f5ae23172ba60fd7bc11ef671cca678842c7072/LICENSE (Apache-2.0); https://huggingface.co/depth-anything/Video-Depth-Anything-Small/blob/256875362cff76724b920335dfb4b29dd611f66e/README.md (license: apache-2.0)
Input size
any H x W >= 1, resized or padded inside the model
Tested up to (portable, fp16, RTX 5060 Ti 16 GB): video-depth-anything-metric-s 3000x4000 (1.8 GiB peak allocated, 3.5 reserved); video-depth-anything-s 3000x4000 (1.8 GiB peak allocated, 3.5 reserved)
Speed and memory
Depth, synthetic
| model | input | time (ms): original fp32 | original | portable | kernels | kernels speed-up: vs original | vs portable | peak memory (MiB, allocated / reserved): original | portable | kernels |
|---|---|---|---|---|---|---|---|---|---|---|
video-depth-anything-metric-s |
1080x1920 | β | β | β | 200 | β | β | β | β | 1967 / 2206 |
Depth, TUM RGB-D fr1_room, 1080x1920
| model | time (ms): original | portable | kernels | kernels speed-up: vs original | vs portable | peak memory (MiB, allocated / reserved): original | portable | kernels |
|---|---|---|---|---|---|---|---|---|
video-depth-anything-metric-s |
10776 | 2005 | 1848 | 5.8x | 1.1x | 8264 / 11916 | 8203 / 8848 | 8208 / 8630 |
video-depth-anything-s |
11626 | 1434 | 1284 | 9.1x | 1.1x | 8263 / 11890 | 8201 / 8652 | 8204 / 8782 |
Startup and first call
| model | load (s): original | portable | kernels | first call (s): original | portable | kernels |
|---|---|---|---|---|---|---|
video-depth-anything-metric-s |
3.85 (1080x1920) | 0.51 (1080x1920) | 0.30 (480x640) | 17.09 (1080x1920) | 2.33 (1080x1920) | 0.43 (480x640) |
video-depth-anything-s |
0.26 (1080x1920) | 0.18 (1080x1920) | 0.12β (1080x1920) | 14.04 (1080x1920) | 1.46 (1080x1920) | 1.30 (1080x1920) |
torch.compile and TensorRT
| model | input | time: portable | compile | TensorRT fp16 | kernels | preparation: compile (cold first call) | ONNX + engine build | peak MiB (alloc / reserved): portable | compile | kernels | TensorRT (engine + torch) |
|---|---|---|---|---|---|---|---|---|---|---|---|
video-depth-anything-s |
1080x1920 | 1434 ms | 1314 ms | 524 ms | 1284 ms | 82.7 s | 181 s | 8201 / 8652 | 7762 / 8530 | 8204 / 8782 | 2753 / 3148 |
Fixed-size cells: time (ms) / peak allocated / peak reserved memory (MiB).
Hardware, protocol and caveats: benchmarks (RTX 5060 Ti 16 GB, sm_120). Original = the original code named above at the precision of portable (fp16, or (fp32) where fp16 is unsupported or not finite); original fp32 = its default fp32 run. Peak memory is the torch allocator's.
Accuracy
On public datasets
AbsRel / delta1 after a least-squares scale + shift fit (in disparity space for inverse-depth models), mean over images. Video: one call per sequence, one scale + shift per sequence, AbsRel / delta1 over all its pixels with Kinect depth, mean over sequences. Original = the original code named under Details, same checkpoint and protocol.
| variant | dataset | original fp32 | original fp16 | portable | kernels | note |
|---|---|---|---|---|---|---|
video-depth-anything-s |
TUM RGB-D video (3 sequences, first 110 frames each), AbsRel / delta1 | 0.0549 / 0.9802 | 0.0549 / 0.9802 | 0.0550 / 0.9802 | 0.0549 / 0.9802 | |
video-depth-anything-s |
TUM RGB-D video (3 sequences, first 1000 frames each), AbsRel / delta1 | 0.1120 / 0.8716 | ||||
video-depth-anything-metric-s |
TUM RGB-D video (3 sequences, first 110 frames each), AbsRel / delta1 | 0.0786 / 0.9769 | 0.0785 / 0.9769 | 0.0786 / 0.9769 | 0.0785 / 0.9769 | |
video-depth-anything-metric-s |
TUM RGB-D video (3 sequences, first 1000 frames each), AbsRel / delta1 | 0.1214 / 0.8883 |
Measured on NVIDIA GeForce RTX 5060 Ti, torch 2.14.1+cu130, fastforks a49aaba7, 2026-11-12.
Parity with the original code
Every variant is checked against reference outputs of the original code at its pinned commit. CPU, float32: max |diff| <= 1.3e-05 vs the original. GPU, float16 (the default), portable and kernels: mean |diff| <= 0.42 % and p99 <= 0.55 % of the reference's median |value| (absolute error of the output inverse_depth).
Limitations
model.video(frames)runs one video in time order (each window attends over time and is aligned to the previous one); calling the model with an image or a batch runs each image as a one-frame video.- Every video runs at least one 32-frame window (shorter videos are padded with copies of their last frame): one image costs about as much as 32 frames in the temporal head.
- The whole video's depth is kept on the device as float32 (L x H x W); split very long videos.
Usage
from fastforks import depth
model = depth.load("video-depth-anything-metric-s") # mode="auto": the preferred backend from recorded measurements
result = model("room.jpg")
result.inverse_depth # [H, W] float32 tensor; result.kind says what the maps mean
# models that take a camera: model("room.jpg", intrinsics=K), K [3, 3] in pixels
result.backend # what ran, e.g. Backend(mode='portable', device='cuda:0', dtype='float16')
Install with pip install fastforks (see https://github.com/kleinicke/fastforks). Other variants: video-depth-anything-metric-s, video-depth-anything-s.
Kernels backend
depth.load("video-depth-anything-metric-s", mode="kernels") runs our CUDA kernels (fastbb). It needs a prebuilt kernel for your GPU architecture (downloaded from https://huggingface.co/fastforks/kernels on first use) and a recent NVIDIA driver; if there is none it raises BackendUnavailableError with the reason and never falls back silently. mode="auto" picks kernels when it can run and logs the choice.