Summary and Motivation:

Egocentric vision is gathering huge momentum. One of the objectives of this scheme is to identify, recognise and include recent published works that advance egocentric vision. This will assist in the formation of a solid community as well as increase the profile of the area. While egocentric vision papers are now being selected in top conferences as orals, highlights and even paper awardees, the area is still exploratory in nature. Recognising a selective set of outstanding works (up to 10) every year, and introducing such efforts to the community is one way to highlight the latest and greatest works and will form the main task of the EgoVis Board.

Eligibility Criteria:

• In this first round, we invite peer-reviewed papers (at conferences or journals) that were accepted between Jan 2022 and Dec 2023. We consider the date of formal acceptance (without any revisions), rather than the conference commencement date or date of proceedings. For conferences, this round will consider papers accepted at: CVPR’22, ECCV’22, NeurIPS’22, ICLR’23, CVPR’23, ICCV’23, NeurIPS’23, ICLR’24. This is a non-exclusive list of venues, and other venues in graphics, robotics or HCI that contribute to egocentric vision will also be considered.
• Submissions to journals should be original contributions and are not extensions of a prior conference publication.
• Evidence of paper’s acceptance is captured through an email or through a link to the published proceedings (if already published).
• Papers from both academia and industry will be considered. There are no additional constraints on authorship, and no maximum number of nominations per author.

Process:

We invite self nominations through a form, to be filled by the authors themselves. We invite papers that have already gone through peer-reviewing (i.e. accepted in a major conference or journal of relevance).

The selection process will focus on the original and innovative contribution of the egocentric vision research, as the paper’s prime focus. We strongly encourage papers that also offer open contributions to the community (code, models and data). The board reserves the right to consider distinguished papers that have not been nominated by the authors should they feel a paper deserves to be nominated. The board also reserves the right to recruit reviewers for the assessment of submissions and overall ranking. The reviewing will be single-blind reserving the anonymity of reviewers. Individual board members will recuse themselves from discussions related to papers in which they are a co-author or are otherwise in-conflict with any of the authors through institutions, close collaboration or advisor-student relationships.

The nomination form will request not only the accepted manuscript but also additional context (reviews, meta-reviews, published code, …) to assist with the selection process. Additionally, a prologue (1-pager, any template) by the authors on the paper’s standing, current or potential impact, will be considered. Letters of recommendations from other members of the community are not needed and will not be considered.

Dates:

The scheme will be open for submissions on 20th of Feb 2024, with nominations accepted until 1st of May 2024. Winners will be announced at the 1st EgoVis workshop alongside CVPR on the 17th of June 2024.

Each winning paper will receive a certificate, monetary award of $1000 and a highlight on the EgoVis website and through social media.

Organising Committee:

The organising EgoVis board represents a diverse set of volunteers who have previously contributed to organising workshops and challenges in Egocentric Vision. Their main motivation, as listed above, is to promote and highlight the latest in the field through this scheme.

EgoVis Board Members (alphabetically)
David Crandall - Indiana University
Dima Damen - University of Bristol and Google DeepMind
Giovanni Maria Farinella - University of Catania and Next Vision
Kristen Grauman - Meta and UT Austin
Jitendra Malik - UC Berkeley and Meta
Richard Newcombe - Meta Reality Labs
Marc Pollefeys - Microsoft and ETH Zurich
Yoichi Sato - University of Tokyo

Copyright, EgoVis Board, 2024