Instructions to use depth-anything/Depth-Anything-V2-Large with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- DepthAnythingV2
How to use depth-anything/Depth-Anything-V2-Large with DepthAnythingV2:
# Install from https://github.com/DepthAnything/Depth-Anything-V2 # Load the model and infer depth from an image import cv2 import torch from depth_anything_v2.dpt import DepthAnythingV2 # instantiate the model model = DepthAnythingV2(encoder="vitl", features=256, out_channels=[256, 512, 1024, 1024) # load the weights filepath = hf_hub_download(repo_id="depth-anything/Depth-Anything-V2-Large", filename="depth_anything_v2_vitl.pth", repo_type="model") state_dict = torch.load(filepath, map_location="cpu") model.load_state_dict(state_dict).eval() raw_img = cv2.imread("your/image/path") depth = model.infer_image(raw_img) # HxW raw depth map in numpy - Notebooks
- Google Colab
- Kaggle
Output unit
#3
by hzialo - opened
Hi!
The model returns an array where I estimate each value to be the depth of that pixel. But, in what units? I would like to know it in millimeters, I don't know if it is possible. Best regards!
Do you figured it out ? I also would like to know if we can get the z depth values data based from my input 2D image.
Thanks
@BrioAnimation & @hzialo did you guys figure this out?
We can set up some experiments to test. Hopefully the depth field is not relative.