Voice-based interface for accessible soundscape composition: composing soundscapes by vocally querying online sounds repositories

Abstract

This paper presents an Internet of Audio Things ecosystem devised to support soundscape composition via vocal interactions. The ecosystem involves a commercial voice-based interface and the cloud-based repository of audio content Freesound.org. The user-system interactions are exclusively based on vocal input/outputs, and differ from the conventional methods for retrieval and sound editing which involve a browser and programs running on a desktop PC. The developed ecosystem targets sound designers interested in soundscape composition and in particular the visually-impaired ones, with the aim of making the soundscape composition practice more accessible. We report the results of a user study conducted with twelve participants. Overall, results show that the interface was found usable and was deemed easy to use and to learn. Participants reported to have enjoyed using the system and generally felt that it was effective in supporting their creativity during the process of composing a soundscape.