Workshop on Long-Range Recognition (LRR)

The topic of vision-based recognition in uncontrolled environments has occupied researchers for decades. The addition of large standoff (lateral and/or vertical) distances between sensing platforms and items being sensed adds new challenges to recognition in less-constrained circumstances. There are focused research programs addressing this issue, along with the supporting challenges of data collection, data curation, etc. Strong-performing systems for recognition “at range” and involving multiple modalities (fusing multiple frames in video sequences and multiple modes, e.g. for biometrics including face, body, and gait) will enable use of such systems in a variety of remote monitoring contexts.

The proposed workshop will allow current and new work in this domain to be disseminated. In the process of dissemination, we expect to

• develop implicit consensus for good current practices on data-related matters,

• highlight architectures for multimodal recognition that seem to offer potential for strong performance, and

• identify topics that need new focused attention.

Program

10:45 - Welcome and Overview

10:50 - Invited Speaker: Irfan Essa

11:30 - Accepted Paper Talks (15 minutes each, including Q&A)

• Expanding Accurate Person Recognition to New Altitudes and Ranges: The BRIAR Dataset. Presented by David Bolme (ORNL)

• Long range gait matching using 3D body fitting with gait-specific motion constraints. Presented by Cole Hill (USF)

• Attentive Sensing for Long-Range Face Recognition. Presented by Helio Perroni Filho (York University)

12:15 - 1:45 - Lunch break

1:45 - Invited Speaker: Lars Ericson

2:15 - Take Two Presentations (10 minutes each, including Q&A)

• Long-range Face Recognition - Xiaoming Liu (MSU)

• On the Effect of Atmospheric Turbulence in the Feature Space of Deep Face Recognition - Wes Robbins (UCCS)

• Unsupervised and self-adaptive techniques for cross-domain person re-identification - Gabriel Bertocco (University of Campinas)

• Learning Domain and Pose Invariance for Thermal-to-Visible Face Recognition and Multi-Context Grouped Attention for Unsupervised Person Re-Identification - Ben Riggan (University of Nebraska)

• Recognizing Faces through Atmospheric Turbulence - Vishal Patel (JHU)

3:05 - Afternoon break

3:45 - Forthcoming Datasets and Challenges in Long-range Recognition - short presentations and panel discussion

"Take Two" call for Presentations

Due to COVID and other travel restrictions, many authors have been unable to present work related to the workshop themes at recent conferences. The workshop will feature a session for "take two" presentations, where authors can give a short oral presentation and (pending the availability of poster boards during the day) present a poster for relevant papers that have already been accepted elsewhere. The presentation should be of a recent paper (since WACV 2022) and be technically relevant to the topic of the workshop. Though papers which have not previously been presented in-person may be prioritized, papers that have already been presented in person will also be considered.

"Take Two" submissions will be considered on a rolling basis. In order to plan their potential attendance at the workshop, authors can specify a date by which they require notification.

To be considered, fill out a "Take Two" submission via CMT: https://cmt3.research.microsoft.com/LRR2023/Track/2/Submission/Create

Organizers

General co-chairs:

• Terry Boult, University of Colorado - Colorado Springs

• Rama Chellappa, John's Hopkins University

• Patrick Flynn, Notre Dame

Program co-chairs:

• Scott McCloskey, Kitware

• Vishal Patel, John's Hopkins University

• Ben Riggan, University of Nebraska

Workshop reviewers:

• Deeksha Arun, Notre Dame

• David Lindenbaum, Accenture Federal Services

• Ioannis Kakadiaris, University of Houston

• Mark Keck, STR