Issues

• The demo on huggingface can not usable

  DepthAnything/Video-Depth-Anything#109 · Filter by author HarisIqbal89 opened on Aug 8, 2026

• fisheye camera support?

  DepthAnything/Video-Depth-Anything#108 · Filter by author stihuangyuan opened on Jun 25, 2026

• Stream mode always cache the first frame.

  DepthAnything/Video-Depth-Anything#105 · Filter by author MrLuer opened on Dec 15, 2025

• Feature request

  DepthAnything/Video-Depth-Anything#104 · Filter by author Archit01 opened on Nov 30, 2025

• 🐛 NaN Loss with DINOv3 Backbone During Training

  DepthAnything/Video-Depth-Anything#103 · Filter by author ecchi-cmd opened on Nov 10, 2025

• Does this require nvidia gpu?

  DepthAnything/Video-Depth-Anything#101 · Filter by author Katsuyamaki opened on Oct 23, 2025

• Will Depth Anything 3 address the common issue of missing vertical sharp edges and thin lines?

  DepthAnything/Video-Depth-Anything#100 · Filter by author KolaKater opened on Oct 20, 2025

• --save_exr results in black image, .mp4 work fine

  DepthAnything/Video-Depth-Anything#98 · Filter by author Poppi2 opened on Oct 19, 2025
  1

• Can only estimate for a few seconds

  DepthAnything/Video-Depth-Anything#92 · Filter by author missTL opened on Sep 12, 2025
  5

• Pulling frame-by-frame, pixel-by-pixel depth maps from inference script

  DepthAnything/Video-Depth-Anything#90 · Filter by author sunshineinsandiego opened on Aug 31, 2025

• Question Regarding Sky Handling During Training

  DepthAnything/Video-Depth-Anything#85 · Filter by author rainyNighti opened on Aug 26, 2025

• DINOv3?

  DepthAnything/Video-Depth-Anything#84 · Filter by author noahclements opened on Aug 25, 2025
  1

• performance drop between the streaming model and the offline model? 这个问题解决了吗？ 可以通过保留关键帧的方法消除这个损失吗？不要只使用隐藏是空头。

  DepthAnything/Video-Depth-Anything#83 · Filter by author cudnn opened on Aug 25, 2025

• Does it work for IR image?

  DepthAnything/Video-Depth-Anything#81 · Filter by author suni1015 opened on Aug 21, 2025

• Add a hugging face format version?

  DepthAnything/Video-Depth-Anything#79 · Filter by author lc700x opened on Aug 20, 2025

• How to finetune the model

  DepthAnything/Video-Depth-Anything#75 · Filter by author kyzira opened on Aug 12, 2025
  2

• Multi-GPU inference?

  DepthAnything/Video-Depth-Anything#72 · Filter by author zeroRoman opened on Jul 23, 2025

• "ValueError: Could not find a backend to open"

  DepthAnything/Video-Depth-Anything#64 · Filter by author JamesLAVFX opened on Jun 24, 2025

• code about metric depth prediction

  DepthAnything/Video-Depth-Anything#60 · Filter by author JeffZuo953 opened on Jun 9, 2025

• Get global scale and shift parameters from MoGe to map relative depth to metric depth

  DepthAnything/Video-Depth-Anything#52 · Filter by author hgolestaniii opened on May 14, 2025

• Region on interest

  DepthAnything/Video-Depth-Anything#46 · Filter by author sarthak247 opened on Apr 16, 2025
  1

• how to export onnx model ?

  DepthAnything/Video-Depth-Anything#31 · Filter by author xqh5201314 opened on Feb 22, 2025
  1

• Export onnx dynamic shape input ？

  DepthAnything/Video-Depth-Anything#22 · Filter by author juntaosun opened on Feb 8, 2025
  3

• CoreML conversion?

  enhancement
  DepthAnything/Video-Depth-Anything#18 · Filter by author ckse93 opened on Jan 31, 2025
  1

• Is there any way to implement "enable_model_cpu_offload" or "enable_sequential_cpu_offload" so this can work with lower VRAM cards like a RTX 2060?

  enhancement
  DepthAnything/Video-Depth-Anything#16 · Filter by author redman4585 opened on Jan 29, 2025
  6