Views
No views yet
foundationstereo/ — Fast-FoundationStereo1HuTt7UIp7gQsMiDvJwVuWmKpvFzIIMap.
Each timestamped folder is one complete model (model_best_bp2_serialize.pth + cfg.yaml):| Folder | Notes |
|---|---|
23-36-37 | Most accurate (PyTorch 49.4 ms / TRT 23.4 ms) |
20-26-39 | Middle |
20-30-48 | Fastest (PyTorch 38.4 ms / TRT 16.6 ms) |
15-44-51 | Extra model included in the source folder |
onnx/ holds pre-exported ONNX models for 20_26_39, 20_30_48, 23_36_37
at 320x736 / 576x960 and 4 / 8 iterations.liteanystereo/ — LiteAnyStereo1UvDx296pVk7pC2rozKIpQF_EXcOleZOB.LAS1/ — LAS1 checkpoints, incl. LiteAnyStereo.pth (not available on HF elsewhere).LAS2/ — LAS2 S/M/H/L (same as tomtomtommi/LiteAnyStereoV2, included for completeness).mobilestereonet/ — MobileStereoNet*DS+KITTI2015* and *SF+DS+KITTI2015*
are the best on KITTI2015 val.1# NOTE: on the corporate network, add these two lines BEFORE importing
2# huggingface_hub, or the TLS interception makes SSL verification fail:
3import truststore
4truststore.inject_into_ssl()
5
6from huggingface_hub import snapshot_download, hf_hub_download
7
8# whole repo
9snapshot_download(repo_id="Miayan/stereo-matching-weights", local_dir="./weights")
10
11# single file
12hf_hub_download(repo_id="Miayan/stereo-matching-weights",
13 filename="foundationstereo/23-36-37/model_best_bp2_serialize.pth",
14 local_dir="./weights")