EPIC-KITCHENS-100
EPIC-KITCHENS-100 is the largest annotated egocentric (head-mounted camera) dataset of unscripted daily kitchen activities, comprising 100 hours of video (20M frames) across 700 variable-length videos recorded by 37…
Source notes
EPIC-KITCHENS-100 is the largest annotated egocentric (head-mounted camera) dataset of unscripted daily kitchen activities, comprising 100 hours of video (20M frames) across 700 variable-length videos recorded by 37 participants in 45 kitchens across 4 cities. It provides ~90K fine-grained action segments with dense language narrations, plus optical flow, audio, object segmentation masks, and hand-object bounding boxes. It is widely used as a human-demonstration resource for egocentric perception, action recognition/anticipation, and embodied/robot learning research.
- Scale
- 100 hours
- Formats
- custom
- License
- CC-BY-NC-4.0
- Published
- 2022-06-01
Decision summary
human_demo
Metadata requires review against the official source before publication.
Read the dataset manifest and feature schema.
Release history
Source-backed release timing for this canonical dataset record.
- Release evidence
Dataset series published
Release timing is recorded from official dataset metadata.
Selection readiness
6 evidence dimensions for deciding whether this dataset is ready to inspect, compare, or adopt. This is not a model benchmark.
Catalog evidence · not task performance
56/100
Provisional
3/6 dimensions scored · 41% confidence
Read the evidence behind all 6 dimensions
Access and governance
The declared license restricts commercial use.
Schema and signal coverage
No machine-readable schema has been verified yet.
Policy training readiness
Observation/action alignment has not been established.
World-model readiness
Temporal observations plus geometry or semantic context are declared; sample alignment remains to be audited.
Failure and recovery readiness
No verified failure/recovery annotation evidence is available yet.
Download and processing readiness
Scale is declared; transfer and processing estimates are not measured.
Review unresolved evidence and next checks
Signal gaps
- Gaze / attention. Not enough evidence is available to classify this signal.
- Feedback / correction / failure. Not enough evidence is available to classify this signal.
- Sim-real pairing. Not enough evidence is available to classify this signal.
Next checks
- Read the dataset manifest and feature schema.
- Run a bounded sample audit before assigning Strong readiness.
Dataset facts
- Source
- University of Bristol (with University of Catania and University of Toronto)
- Evidence
- secondary claim
- Formats
- custom
- episodes
- 700
- hours
- 100
- size
- 796 GB
Loop signals
4/7 present or partial
No decision-grade evidence captured yet.
unknownNo decision-grade evidence captured yet.
unknownView 5 more signal categories
video · EPIC-KITCHENS-100 is the largest annotated egocentric (head-mounted camera) dataset of unscripted daily kitchen activities, comprising 100 hours of video (20M frames) across 700 variable-length videos recorded by 37 participants in 45 kitchens across 4 cities. It provides ~90K fine-grained action segments with dense language narrations, plus optical flow, audio, object segmentation masks, and hand-object bounding boxes. It is widely used as a human-demonstration resource for egocentric perception, action recognition/anticipation, and embodied/robot learning research.
presentEPIC-KITCHENS-100 is the largest annotated egocentric (head-mounted camera) dataset of unscripted daily kitchen activities, comprising 100 hours of video (20M frames) across 700 variable-length videos recorded by 37 participants in 45 kitchens across 4 cities. It provides ~90K fine-grained action segments with dense language narrations, plus optical flow, audio, object segmentation masks, and hand-object bounding boxes. It is widely used as a human-demonstration resource for egocentric perception, action recognition/anticipation, and embodied/robot learning research.
partiallanguage · EPIC-KITCHENS-100 is the largest annotated egocentric (head-mounted camera) dataset of unscripted daily kitchen activities, comprising 100 hours of video (20M frames) across 700 variable-length videos recorded by 37 participants in 45 kitchens across 4 cities. It provides ~90K fine-grained action segments with dense language narrations, plus optical flow, audio, object segmentation masks, and hand-object bounding boxes. It is widely used as a human-demonstration resource for egocentric perception, action recognition/anticipation, and embodied/robot learning research.
presentNo decision-grade evidence captured yet.
unknownOpen · CC-BY-NC-4.0 · custom
partialEvidence details and provenance
Official signal claims
Schema and annotations
No machine-readable schema facts are captured.
Sample verification
Pending. Metadata does not prove sample coverage, alignment, or file integrity.
curated official source
Integration notes
- Metadata requires review against the official source before publication.