A grid of RGB fisheye images, each part of the Project Aria Pilot Dataset.

PROJECT ARIA OPEN DATASETS

Research datasets for accelerating state-of-the-art

AR and AI technologies will have a profound impact on the world.

Datasets captured using Project Aria allow us to explore the technical and societal challenges associated with future AR and AI devices, prior to those devices being publicly available.

By making Project Aria datasets available to academic researchers, we hope we can support the growth of AI and ML, enabling us to help build the future for the better, together.

ACCESS

Project Aria Open Datasets

A thumbnail shows a snapshot from the Aria Everyday Activities Dataset

Aria Everyday Activities
(Re-release of Aria Pilot Dataset)

A real-world dataset, consisting of common activities within common environments

A thumbnail shows a snapshot from the Aria Synthetic Environments Dataset

Aria Synthetic Environments

A procedurally-generated synthetic dataset for large-scale ML training

A thumbnail shows a snapshot from the Aria Digital Twin Dataset

Aria Digital Twin

A real-world dataset, with dynamic and photorealistic digital counterpart

Challenges

Open research challenges to help establish baselines for machine perception and AI capabilities.

Tools

Open source models, tools and libraries help to provide researchers the means to work with Project Aria data, and conduct research responsibly.

Paper

Read the technical paper on arXiv to learn more about the Project Aria glasses specification.

Documentation

Learn more about Project Aria open tools by diving into the documentation on GitHub

Subscribe to Project Aria Updates

Stay in the loop with the latest datasets and challenges from Project Aria.

SIGN UP FOR EMAIL UPDATES