The Current Landscape of Audio Description Research A Survey

Labdhi Gada, Ankit Rajiv Jindal, Neelam Jain

Abstract: Audio description (AD) enables those with visual difficulties to hear what cannot be seen on film, television, and real-time footage. Visually impaired persons could benefit from AD by receiving an audio rendition of the visual content. The task of automatically summarising or subtitling a natural language video is considered one of the most challenging computer vision issues to solve. However, the problem is not minor, especially when a film has several notable events, which often happen in real recordings. This research project aims at providing dense subtitles, which include both - the identification and description of events in a video.

Keywords: Audio description, captioning, audio captioning, screen captioning, media accessibility