Datasets › EGTEA

EGTEA (EGTEA Gaze+)

Introduced by Yin Li et al. in In the Eye of Beholder: Joint Learning of Gaze and Actions in First Person Video1 Sep 2018 archive 2025-07-28

Extended GTEA Gaze+ EGTEA Gaze+ is a large-scale dataset for FPV actions and gaze. It subsumes GTEA Gaze+ and comes with HD videos (1280x960), audios, gaze tracking data, frame-level action annotations, and pixel-level hand masks at sampled frames. Specifically, EGTEA Gaze+ contains 28 hours (de-identified) of cooking activities from 86 unique sessions of 32 subjects. These videos come with audios and gaze tracking (30Hz). We have further provided human annotations of actions (human-object interactions) and hand masks.

The action annotations include 10325 instances of fine-grained actions, such as "Cut bell pepper" or "Pour condiment (from) condiment container into salad".

The hand annotations consist of 15,176 hand masks from 13,847 frames from the videos.

Source: http://cbs.ic.gatech.edu/fpv/ Image Source: http://cbs.ic.gatech.edu/fpv/

Homepage Benchmarks Papers JSON

Benchmarks archive 2025-07-28

All 3 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

		First row (archive order)	Paper	Code
Egocentric Activity Recognition	EGTEA	LaViLa (Finetuned, TimeSformer-L) Average Accuracy 81.75	Learning Video Representations from Large Language Models	facebookresearch/lavila +2	6	Compare
Action Anticipation	EGTEA	UADT Top-1 Accuracy 68.4	Uncertainty-aware Action Decoupling Transformer for...	—	3	Compare
Long-tail Learning	EGTEA	CDB-loss (3D- ResNeXt101) Average Precision 63.86	Class-Wise Difficulty-Balanced Loss for Solving Class-Imbalance	hitachi-rd-cv/CDB-loss	3	Compare

Papers archive 2025-07-28

12 shown of 12 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 100. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

			Date	Samples run Syntology
Uncertainty-aware Action Decoupling Transformer for Action Anticipation	0	1	1 Jan 2024	not harvested
Learning Video Representations from Large Language Models	3	1	8 Dec 2022	ran 6 of 20 samples (14 unverified; 20 pointer-only for licence)
Interaction Region Visual Transformer for Egocentric Action Anticipation	1	1	25 Nov 2022	not harvested
Predicting the Next Action by Modeling the Abstract Goal	0	1	12 Sep 2022	not harvested
Group Contextualization for Video Recognition	1	1	18 Mar 2022	ran 5 of 5 samples (0 unverified)
Integrating Human Gaze into Attention for Egocentric Activity Recognition	1	1	8 Nov 2020	not harvested
Class-Wise Difficulty-Balanced Loss for Solving Class-Imbalance	1	1	5 Oct 2020	ran 0 of 2 samples (2 unverified; 2 pointer-only for licence)
Symbiotic Attention with Privileged Information for Egocentric Action Recognition	0	1	8 Feb 2020	not harvested
Class-Balanced Loss Based on Effective Number of Samples	11	1	16 Jan 2019	ran 7 of 27 samples (20 unverified)
LSTA: Long Short-Term Attention for Egocentric Action Recognition	1	1	26 Nov 2018	not harvested
Attention is All We Need: Nailing Down Object-centric Attention for Egocentric Activity Recognition	1	1	31 Jul 2018	not harvested
Focal Loss for Dense Object Detection	234	1	7 Aug 2017	ran 11 of 11 samples (0 unverified; 6 pointer-only for licence)

Dataset loaders archive 2025-07-28

No loader listed in the archive.

Tasks archive 2025-07-28

Long-tail Learning Action Anticipation Egocentric Activity Recognition

License archive 2025-07-28

No licence recorded in the archive. Absence here is not a statement about the dataset's terms.

Modalities archive 2025-07-28

No modality tagged.

Languages archive 2025-07-28

No language tagged.

Variants archive 2025-07-28

No variants listed.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections