Frequently asked questions

WalkSentry is a student research lab working on pedestrian safety with computer vision. We focus on two groups: blind and low-vision (BLV) people, and pedestrians walking with divided attention, for example while using their phones.

Our members are at York University in Toronto and Amirkabir University of Technology in Tehran. We work together remotely and collect data in both cities.

The lab is supervised by Hamed Farbeh of Amirkabir University of Technology. You can meet the full team on the About Us page.

We work on three connected areas: obstacle-detection datasets captured from a pedestrian's point of view, real-time object detectors small enough to run on edge devices such as phones, and assistive navigation systems built on top of them.

ROD-Dataset is our pedestrian-perspective benchmark for obstacle detection. It has more than 24,000 eye-level images and over 40,000 human-verified annotations across 25 obstacle classes, from vehicles and people to stairs, manholes and traffic cones. The paper is currently under review.

We plan to release datasets and code alongside our papers. Release links and license details will appear on the Articles & Links page and on our GitHub organization as they become available.

Each paper on the Articles & Links page has a Copy BibTeX button. Some papers are still under review, so their citation details may change once they are published.

Not yet. The app is in development. Download links will appear on the App & Demo page as soon as it is released.

The app is designed for blind and low-vision pedestrians and for anyone who walks with divided attention. Accessibility is a core requirement of everything we build, not an add-on.

We welcome students, researchers and organizations working with the blind and low-vision community. The easiest way to reach us is through our LinkedIn page. Tell us a little about yourself and how you would like to be involved.