Datasets:
Tasks:
Object Detection
Modalities:
Image
Languages:
English
Size:
100K<n<1M
Tags:
Multi-Object-Tracking
License:
Download annotations/hard_tracks_dataset_coco_val.json from mscheidl/htd: direct link, hf CLI and curl.
- Browser
- Download file 79.8 MB
-
https://huggingface.co/datasets/mscheidl/htd/resolve/main/annotations/hard_tracks_dataset_coco_val.json
- Command line
-
hf download hf://datasets/mscheidl/htd/annotations/hard_tracks_dataset_coco_val.json
-
curl -L -o hard_tracks_dataset_coco_val.json https://huggingface.co/datasets/mscheidl/htd/resolve/main/annotations/hard_tracks_dataset_coco_val.json
79.8 MB
- Xet hash:
- cc5142746650fa0ecefe00651c3a300ebd05a258af6c6903c79a20e4d308b942
- Size of remote file:
- 79.8 MB
- SHA256:
- 4561050b4daf7e0800a4e14165dbb189d852fa90ac424677c89c5424d275c975
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.