More about HKUST
AI-Powered Interactive Systems that Improve Video Accessibility for Blind and Low Vision Viewers
PhD Qualifying Examination
Title: "AI-Powered Interactive Systems that Improve Video Accessibility for
Blind and Low Vision Viewers"
by
Mr. Shuchang XU
Abstract:
Video has become a central medium for communication, education, and
entertainment, yet it remains largely inaccessible to blind and low vision
(BLV) viewers, who often lack access to the visual information essential for
understanding videos. The primary approach for making videos accessible is
audio description (AD), which provides spoken narration of key visual
elements alongside the video. However, current AD practices face two
primary challenges: the labor-intensive process of creating AD and the lack
of interactivity to accommodate diverse viewer needs. This survey presents a
comprehensive review of AI-powered interactive systems developed to address
these challenges. We classify prior work into two main categories: (1)
audio description creation, which leverages AI to enhance the efficiency
and scalability of AD creation; and (2) viewer interaction, which enables
BLV viewers to explore video content according to their individual
information needs. Using this taxonomy, we analyze representative systems,
identify emerging research trends, and discuss future directions toward more
coherent, context-aware, and immersive video experiences for BLV audiences.
Date: Wednesday, 26 November 2025
Time: 9:00am - 11:00am
Venue: Room 4472
Lifts 25/26
Committee Members: Prof. Huamin Qu (Supervisor)
Dr. Xiaojuan Ma (Chairperson)
Dr. Anyi Rao (AMC)