ISCA Archive Interspeech 2025
ISCA Archive Interspeech 2025

Concurrent Speech and Auditory Tag Clouds for Non-Visual Web Interaction

Dhia Eddine Merzougui, Nilesh Tete, Fabrice Maurel, Gaël Dias, Mohammed Hasanuzzaman, Aurélien Bournonville, Edgar Madelaine, Thomas Berthelin Le Tellier, François Ledoyen, Laure Poutrain-Lejeune, François Rioult, Jérémie Pantin

Millions of visually impaired individuals face significant barriers to independent information access, limiting their autonomy and self-determination. Despite advancements in assistive technologies, challenges remain, particularly in handling web document structure and multi-channel information transposition. This paper addresses these issues by introducing the TagThunder experimental framework, which applies the "cocktail party effect" metaphor to transpose the morpho-dispositional semantics of web documents into an auditory tag cloud, enabling rapid non-visual skimming. Additionally, it explores interactive stimuli for structured information scanning through discrete and continuous guiding exploration strategies.