Enabling Eye Tracking for Crowd-Sourced Data Collection With Project Aria

Through a novel, multi-sensor platform in a wearable glasses form factor, paired with crowd-sourced data collection, we have enabled the collection of tens of thousands of egocentric gaze recordings across hundreds of daily tasks without the limitations posed by traditional eye-tracking or lab resea...

Full description

Saved in:
Bibliographic Details
Main Authors: Yusuf Mansour, Ajoy Savio Fernandes, Kiran Somasundaram, Tarek Hefny, Mahsa Shakeri, Oleg V. Komogortsev, Abhishek Sharma, Michael J. Proulx
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11052289/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850117964041289728
author Yusuf Mansour
Ajoy Savio Fernandes
Kiran Somasundaram
Tarek Hefny
Mahsa Shakeri
Oleg V. Komogortsev
Abhishek Sharma
Michael J. Proulx
author_facet Yusuf Mansour
Ajoy Savio Fernandes
Kiran Somasundaram
Tarek Hefny
Mahsa Shakeri
Oleg V. Komogortsev
Abhishek Sharma
Michael J. Proulx
author_sort Yusuf Mansour
collection DOAJ
description Through a novel, multi-sensor platform in a wearable glasses form factor, paired with crowd-sourced data collection, we have enabled the collection of tens of thousands of egocentric gaze recordings across hundreds of daily tasks without the limitations posed by traditional eye-tracking or lab research. The eye tracking model that we describe and open source here was created with a large study of 1500 participants performing eye tracking calibration. Data collection utilized the device sensor suite, allowing us to leverage scene and contextual information across various environments and tasks. This paper will discuss the systems in place and technical steps we have taken to enable eye tracking for the project, and select lessons learned in the process. This large-scale data collection tool and open source eye tracking model will enable advances in mobile and pervasive eye tracking to understand nuances in eye movements and various future AI and eye-tracking applications.
format Article
id doaj-art-1f5b85fcb8034004b86ec9c14e4aed8f
institution OA Journals
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-1f5b85fcb8034004b86ec9c14e4aed8f2025-08-20T02:35:59ZengIEEEIEEE Access2169-35362025-01-011311473611474510.1109/ACCESS.2025.358362311052289Enabling Eye Tracking for Crowd-Sourced Data Collection With Project AriaYusuf Mansour0Ajoy Savio Fernandes1Kiran Somasundaram2https://orcid.org/0000-0001-8554-9083Tarek Hefny3Mahsa Shakeri4Oleg V. Komogortsev5Abhishek Sharma6Michael J. Proulx7https://orcid.org/0000-0003-4066-3645Meta Reality Labs Research, Redmond, WA, USAMeta Reality Labs Research, Redmond, WA, USAMeta Reality Labs Research, Redmond, WA, USAMeta Reality Labs Research, Redmond, WA, USAMeta Reality Labs, Redmond, WA, USAMeta Reality Labs Research, Redmond, WA, USAMeta Reality Labs Research, Redmond, WA, USAMeta Reality Labs Research, Redmond, WA, USAThrough a novel, multi-sensor platform in a wearable glasses form factor, paired with crowd-sourced data collection, we have enabled the collection of tens of thousands of egocentric gaze recordings across hundreds of daily tasks without the limitations posed by traditional eye-tracking or lab research. The eye tracking model that we describe and open source here was created with a large study of 1500 participants performing eye tracking calibration. Data collection utilized the device sensor suite, allowing us to leverage scene and contextual information across various environments and tasks. This paper will discuss the systems in place and technical steps we have taken to enable eye tracking for the project, and select lessons learned in the process. This large-scale data collection tool and open source eye tracking model will enable advances in mobile and pervasive eye tracking to understand nuances in eye movements and various future AI and eye-tracking applications.https://ieeexplore.ieee.org/document/11052289/Contextual AIegocentric perceptioneye trackinghuman-computer interactionmachine learningpervasive computing
spellingShingle Yusuf Mansour
Ajoy Savio Fernandes
Kiran Somasundaram
Tarek Hefny
Mahsa Shakeri
Oleg V. Komogortsev
Abhishek Sharma
Michael J. Proulx
Enabling Eye Tracking for Crowd-Sourced Data Collection With Project Aria
IEEE Access
Contextual AI
egocentric perception
eye tracking
human-computer interaction
machine learning
pervasive computing
title Enabling Eye Tracking for Crowd-Sourced Data Collection With Project Aria
title_full Enabling Eye Tracking for Crowd-Sourced Data Collection With Project Aria
title_fullStr Enabling Eye Tracking for Crowd-Sourced Data Collection With Project Aria
title_full_unstemmed Enabling Eye Tracking for Crowd-Sourced Data Collection With Project Aria
title_short Enabling Eye Tracking for Crowd-Sourced Data Collection With Project Aria
title_sort enabling eye tracking for crowd sourced data collection with project aria
topic Contextual AI
egocentric perception
eye tracking
human-computer interaction
machine learning
pervasive computing
url https://ieeexplore.ieee.org/document/11052289/
work_keys_str_mv AT yusufmansour enablingeyetrackingforcrowdsourceddatacollectionwithprojectaria
AT ajoysaviofernandes enablingeyetrackingforcrowdsourceddatacollectionwithprojectaria
AT kiransomasundaram enablingeyetrackingforcrowdsourceddatacollectionwithprojectaria
AT tarekhefny enablingeyetrackingforcrowdsourceddatacollectionwithprojectaria
AT mahsashakeri enablingeyetrackingforcrowdsourceddatacollectionwithprojectaria
AT olegvkomogortsev enablingeyetrackingforcrowdsourceddatacollectionwithprojectaria
AT abhisheksharma enablingeyetrackingforcrowdsourceddatacollectionwithprojectaria
AT michaeljproulx enablingeyetrackingforcrowdsourceddatacollectionwithprojectaria