TAO Challenge

TAO is a federated dataset for Tracking Any Object, containing 2,907 high resolution videos, captured in diverse environments, which are half a minute long on average. We adopt a bottom-up approach for discovering a large vocabulary of 833 categories, an order of magnitude more than prior tracking benchmarks. This challenge evaluates multi-object tracking on the 488 categories in TAO which overlap with the LVIS (v0.5) dataset. For questions regarding this challenge, please email tao@motchallenge.net.

Training Set

Sample	Name	FPS	Resolution	Length	Tracks	Boxes	Density	Description	Source	Ref.
	TAO_val	30	1280x720	1460666 (31:29)	5485	113112	0.1	Validation set, for tuning hyperparameters	link	[1]
	TAO_train	30	1280x720	764526 (04:44)	2647	54639	0.1	Training set, for training trackers	link	[1]
	Total			2225192 frm. (74173 s.)	8132	167751	0.1

Test Set

Sample	Name	FPS	Resolution	Length	Tracks	Boxes	Density	Description	Source	Ref.
	TAO_test	30	1280x720	2221846 (34:22)	7972	164650	0.1	Test set, for benchmarking	link	[1]
	Total			2221846 frm. (74062 s.)	7972	164650	0.1

Download