Datasets:
Tasks:
Object Detection
Modalities:
Image
Languages:
English
Size:
100K<n<1M
Tags:
Multi-Object-Tracking
License:
Download annotations/hard_tracks_dataset_coco_test.json from mscheidl/htd: direct link, hf CLI and curl.
- Browser
- Download file 97.8 MB
-
https://huggingface.co/datasets/mscheidl/htd/resolve/main/annotations/hard_tracks_dataset_coco_test.json
- Command line
-
hf download hf://datasets/mscheidl/htd/annotations/hard_tracks_dataset_coco_test.json
-
curl -L -o hard_tracks_dataset_coco_test.json https://huggingface.co/datasets/mscheidl/htd/resolve/main/annotations/hard_tracks_dataset_coco_test.json
97.8 MB
- Xet hash:
- 49621e15339b76d59a8c6893f39b7b627f05c4db7917fe2890adc91e8259df5d
- Size of remote file:
- 97.8 MB
- SHA256:
- 1056f5cfe4688461adfd664801ca6909007184e51da91f11cecf8b206a779a43
·
Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.