Home » Store Front Accessibility – Doorfront.org

Attribution-NonCommercial-ShareAlike 4.0 International

Except where otherwise noted, content on this site is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International license.

Store Front Accessibility – Doorfront.org

People who are blind or have low vision (BLV) still struggle to locate store entrances due to missing geospatial information in current map services. To address this, we developed an AI-enabled crowdsourcing platform for collecting storefront accessibility data, which improves data collection and user engagement. Key features include a gamified credit ranking system, a volunteer contribution estimator, an AI-based pre-labeling function, and an image gallery feature. We integrate the Multi-stage Context Learning and Utilization (MultiCLU) for Storefront Accessibility Detection and Evaluation. A MultiCLU deep learning model and an online machine learning mechanism to iteratively train it with new data. Interviews with six blind adults confirmed that the data collected by DoorFront would significantly benefit their daily travel.

Website:

  • https://doorfront.org
  • Note: We provide a community service letter for anyone who labels and validates storefront accessibility data via Doorfront.org. Please email [email protected] to request the letter.

Publications:

  • Tyler Ortiz, Vicky Tang, Karla Sutton User-Centric Crowdsourcing Approach to Improve Urban Accessibility Data Collection, IEEE MIT Undergraduate Research Technology Conference, MA, Oct. 2024
  • Xuan Wang, Jiawei Liu, Hao Tang, Zhigang Zhu, and William H. Seiple, An AI-enabled Annotation Platform for Storefront Accessibility and Localization, Journal on Technology & Persons with Disabilities. Volume 11, June 2023, Robles, A., (Eds): CSUN Assistive Technology Conference © 2023 California State University, Northridge, pp. 76-94.
  • Xuan Wang, Hao Tang & Zhigang Zhu, A General Context Learning and Reasoning Framework for Object Detection in Urban Scenes. VISAPP 2023, the 18th International Conference on Computer Vision Theory and Applications, Feb 19-21, 2023 full oral paper). [DOI]
  • X. Wang, J. Chen, H. Tang and Z. Zhu. MultiCLU: Multi-stage Context Learning and Utilization for Storefront Accessibility Detection and Evaluation. ACM International Conference on Multimedia Retrieval, Newark, NJ, USA, June 27-30, 2022. Pages 304–312. [doi with supplemental video]
  • Jiawei Liu, Hao Tang, William Seiple, Zhigang Zhu. Annotating Storefront Accessibility Data Using Crowdsourcing. Journal on Technology & Persons with Disabilities. Volume 10, June 2022, pp 154-170.