Dataset profile

HD-EPIC

HD-EPIC extends the EPIC-KITCHENS line with detailed and interconnected annotations for unscripted recipe preparation. It is useful for fine-grained task decomposition and grounded procedural reasoning.

Source institutionUniversity of Bristol and EPIC-KITCHENS collaborators
EgocentricActivities of Daily Living

Overview

HD-EPIC extends the EPIC-KITCHENS line with detailed and interconnected annotations for unscripted recipe preparation. It is useful for fine-grained task decomposition and grounded procedural reasoning.

What's included

  • First-person kitchen recordings
  • Highly detailed action and object annotations
  • Recipe and digital-twin resources

Tasks / activities

recipe preparationfine-grained actionsobject trackinggaze priming

Modalities

Egocentric VideoRGBLanguageAnnotationsPose

Collection methodology

Participants recorded multi-day kitchen activity with detailed annotations grounded in recipes and scene information.

Potential applications

  • Fine-grained procedural understanding
  • Task graph learning
  • Egocentric VQA
  • Object-state tracking

Access & licensing

This is a third-party public/research dataset. We do not own or distribute this dataset. Access and usage are governed by the original publisher's terms.

Access and use are governed by the original HD-EPIC and EPIC-KITCHENS terms.

Limitations / opportunities for custom data

Healthcare workflows require professional tasks, care environments, and interaction constraints outside kitchen recipe preparation.

Need data beyond this benchmark?

We design custom healthcare and caregiving collections around the tasks, experts, environments, and modalities your model requires.

Request Similar Data