3rd Workshop on
Computer Vision Applications
for the Visually Impaired (CVAVI 10)

San Francisco, California, June 14, 2010

A Satellite Workshop of IEEE CVPR 2010

Keynote Speakers

David A. Ross, M.S.E.E., M.Ed.
Atlanta VA Rehab R&D Center of
Excellence in Vision Loss and Aging


John Brabyn, Ph.D.
Director, Rehabilitation Engineering
Research Center, Smith-Kettlewell
Eye Research Inst., San Francisco

NEW: Here are directions to the Hyatt Regency, the conference hotel. The workshop will be held in the Pacific Concourse O, in the basement of the hotel.

Jump to the Program

Background

A growing number of computer vision researchers are becoming interested in applications for persons with visual impairments (VI), including low vision and blindness. Recent advances in algorithms, sensors and embedded computing hold the promise to enable computer vision technology that can address the needs of this disadvantaged population. In particular, some popular smartphones now come equipped with screen readers for accessibility to VI users, high resolution cameras with focusing capabilities, and fairly powerful embedded computers, and thus may serve as an ideal platform for many computer vision-enabled VI tools.

Developing assistive tools for the VI community poses special challenges to the design of appropriate computer vision algorithms. Unlike traditional computer vision applications, in which images are often composed by a normally sighted photographer or acquired by a rigidly mounted camera, images acquired in VI applications are highly unconstrained: a blind user may need to pan a camera (which may be a wearable camera fastened to clothing) to find a desired object of interest, with little prior knowledge of where to aim the camera. Moreover, the user may have little or no awareness of lighting conditions, the presence of occluding objects, and the quality of the images being acquired, which may be out of focus, poorly framed, and motion-blurred. Indeed, VI applications often demand “vision without vision” – the ability of an assistive tool to direct the search of a user with no useful vision towards his or her targets. An added constraint is that the necessary computer vision algorithms must execute in real time (often running on an embedded computing system) to be useful.

In addition, research on applications for the VI population also requires informed knowledge of many human factor issues, including the actual problems impairing the daily quality of life of these users, an awareness of existing assistive aids, and a realistic understanding of the possibilities offered by the available interfaces (visual, audio, tactile). Finally, researchers must adopt a system-level approach to the design of algorithms and hardware that takes into consideration practical factors such as form factor, user interface, size, speed and cost.

This workshop aims to bring together computer vision researchers and experts in VI rehabilitation and assistive technology. Attendance by VI experts outside the computer vision community will also be encouraged. By addressing the context in which assistive technology is used and designed, the workshop seeks to help researchers identify fruitful areas of overlap between the most pressing needs of the VI population and the capabilities of computer vision technology. A Panel Discussion session will facilitate the exchange of ideas between computer vision researchers and VI experts.

Previous versions of this workshop under the same title took place in 2005 and in 2008.

Welcome
8:50-9:10

Session 1:  Technology for Education

J. Black and D. Hayden
The Note-Taker: An Assistive Technology That Allows Students Who Are Legally Blind to Take Notes in the Classroom 
9:10-9:35

B. Fang, F. Oliveira and F. Quek
Using Vision-Based Tracking to Support Real-Time Graphical Instruction for Students Who Have Visual Impairments 
9:35-10:00

Coffee break
10:00-10:30

Session 2:  Mobility

V. Pradeep, G. Medioni and J. Weiland
Robot Vision for the Visually Impaired
10:30-10:55

J. Liu, C. Phillips and K. Daniilidis
Video-Based Localization Without 3D Mapping for the Visually Impaired
10:55-11:20

Keynote Talk #1
David Ross
The Human Side of Machine Vision — Divided Attention Issues and the Need for Closed-Loop Control 
11:20-12:05

Demo
V. Kulyukin and A. Kutiyanawala
ShopMobile II: Eyes-Free Supermarket Grocery Shopping for Visually Impaired Mobile Phone Users
12:05-12:30

Lunch
12:30-2:00

Keynote Talk #2
John Brabyn
A Lifetime of Mistakes in Assistive Technology: Lessons for Computer Vision
2:00-2:30

Panel Discussion
2:30-2:50

Session 3:  Annotation/Enhancement

A. Choudhury and G. Medioni
Color Contrast Enhancement for Visually Impaired People 
2:50-3:15

L. Gagnon, C. Chapdelaine, D. Byrns, S. Foucher, M. Héritier and V. Gupta
Computer Vision-Assisted System for Videodescription Scripting
3:15-3:40

Coffee break
3:40-4:00

Session 4:  Object Recognition

E. Christiansen, T. Winlock and S. Belongie
Toward Real-Time Grocery Detection for the Visually Impaired 
4:00-4:25

X. Yang and Y. Tian
Robust Door Detection in Unfamiliar Environments by Combining Edge and Corner Features
4:25-4:50

C. Jayant and J. Bigham
VizWiz::LocateIt — Enabling Blind People to Locate Objects in Their Environment
4:50-5:15

J. Sudol, O. Dialameh, C. Blanchard and T. Dorcey
LookTel — A Comprehensive Platform for Computer-Aided Visual Assistance
5:15-5:40

Demo
V.F. Pamploma. A. Mohan, M. NcKenna, C. Ekeji, M.M. Oliveira and R. Raskar
PerfectSight: Increasing Global Accessibility to Diagnostic Services for Eye Care
5:40-6:00