Home·Curated Catalog·Multimodal
🔀 Curated Catalog · Multimodal

Ego4D

3,670 hours of first-person video from 931 participants in 9 countries.

LQS 88 · gold ⚠ Research-only 3.7K hours of egocentric video 7.2 TB MP4 · JSON Released 2022
Browse commercial Multimodal → Visit original source ↗
Source: ego4d-data.org · maintained by Meta AI + 14 university consortium
3.7K
hours of egocentric video
7.2 TB
Size on disk
88
LQS · gold
2022
First released

About this dataset

Ego4D is a massively multi-institutional first-person video dataset led by Meta FAIR. 3,670 hours of egocentric video from 931 camera wearers in 9 countries, annotated for 5 benchmark tasks: episodic memory, hand-object interaction, AV diarization, social interaction, forecasting. Access requires signing a data use agreement.

Formats
MP4 · JSON

LabelSets Quality Score

LQS is our 7-dimension quality score, computed from the dataset's published statistics. See methodology →

88
out of 100
gold tier

High-quality dataset across most dimensions

Composite score computed from the 7 dimensions below: completeness, uniqueness, validation health, size adequacy, format compliance, label density, and class balance.

Completeness 95
No public completeness metric; using prior for 'expert_curated' datasets.
Uniqueness 95
Manually vetted for uniqueness by maintainer.
Validation 92
Labels produced by domain experts or trained annotators.
Size adequacy 92
3,670 hours — exceeds 100,000 adequacy target for Multimodal.
Format compliance 95
Industry-standard format — drop-in compatible with mainstream tooling.
Label density 52
Average 1.0 labels per item (sparse).
Class balance 75
Moderate class skew — realistic production distribution.

What it's used for

Common tasks and benchmarks where Ego4D is the default or competitive choice.

Sample statistics

What's actually in the dataset — from the maintainer's published stats.

3,670 hours / 931 wearers across 9 countries (India, US, Japan, UK, etc). Head-mounted camera. Multiple benchmark splits.

License

Ego4D is distributed under Ego4D License (DUA required). This is a third-party public dataset; LabelSets indexes and scores it but does not host or redistribute the data. Always verify current license terms with the maintainer before commercial use.

Heads up: this dataset's license restricts commercial use. If you need multimodal data for production, check LabelSets' paid datasets below — every listing has an explicit commercial license.

Need commercial-licensed Multimodal data?

LabelSets sellers offer paid multimodal datasets with what public datasets often can't give you:

Browse paid Multimodal → Sell your dataset

Similar public datasets

Other entries in the Multimodal catalog.

Frequently Asked Questions

Ego4D is distributed under Ego4D License (DUA required), which restricts commercial use. For a commercially-licensed alternative in multimodal, see LabelSets' paid datasets.
Ego4D contains 3,670 hours of egocentric video. 3,670 hours / 931 wearers across 9 countries (India, US, Japan, UK, etc). Head-mounted camera. Multiple benchmark splits.
Ego4D is maintained by Meta AI + 14 university consortium and is available at https://ego4d-data.org/#download. LabelSets indexes and scores this dataset for discoverability but does not redistribute it.
LQS is a 7-dimension quality score (completeness, uniqueness, validation, size adequacy, format compliance, label density, class balance) computed from the dataset's published statistics. Composite scores map to tiers: platinum (≥90), gold (≥75), silver (≥60), bronze (<60). Read the full methodology.