RAVI - Reading Assistant for the Visually Impaired
Updates
  • 16-August-2018: Job opening | full-time | Walk-in: September 4, 2018 at IIT Delhi Click here for advertisement
  • 10-August-2018: As a first-step towards improving reading accessibility of textbooks, over 120 NCERT Hindi books converted from font encoding to Unicode @ CIET, NCERT in association with Saksham and Smart Solutions.

RAVI : Reading Assistant for the Visually Impaired

ABOUT

As technology creates more and more digital data, accessing, analyzing and interpreting this data in various forms has been a major challenge for visually impaired. Volume, cost and effort involved in printing Braille or creating accessible tactile diagrams rules them out except for relatively static content like text books. Though audio using various screen readers have made digital text easily accessible, but they still face significant challenges if they need to access text mixed with mathematical equations, figures and tables. Absence of documents in formats compatible to screen readers create even accessing text in many regional languages a major challenge. Reading Assistant for Visually Impaired (RAVI) focus on making textbooks accessible to readers. We work for eBook accessibility, diagram accessibility, mathematics accessibility as well as accessibility of material in Indian Languages using latest techniques and technologies.

PUBLICATIONS

Template based Approach for Augmenting Image Descriptions

Akshansh Chahal, Manshul Belani, Akashdeep Bansal, Neha Jadhav, and Prof. M. Balakrishnan

Abstract: With the increasing focus on digital learning, it has become extremely important that digital content is available with ease. However, a lot of this digital content is not generated keeping Universal Access in mind. Most of such content available is either completely inaccessible or only partially accessible to the print disabled people.

Abstract: With the increasing focus on digital learning, it has become extremely important that digital content is available with ease. However, a lot of this digital content is not generated keeping Universal Access in mind. Most of such content available is either completely inaccessible or only partially accessible to the print disabled people.One of the major gaps in accessibility of digital content, especially electronic books is the lack of alternative texts for diagrams and ineffective descriptions in cases they are present. The paper discusses the design of a template, which can help in augmenting descriptions for textbook diagrams. The template consists of various components, which are populated using the information present in the diagram or from the text surrounding the diagram in the textbook. This template provides means for generation of comprehensible diagram descriptions, which not only help the user to visualize the diagram but also create a mental model of the layout of the diagram. Observations made during the user study validate the effectiveness of these augmented descriptions.