MeshUp #21 - Deep Image Sonification: Towards a Perceptually Aligned Correspondence between Images and Sounds
This week, Balint Laczko will present Deep Image Sonification: Towards a Perceptually Aligned Correspondence between Images and Sounds.
Abstract
Image sonification—the transformation of data relationships in images into perceived auditory relationships—offers potential for many applications, such as improved medical diagnosis, immersive multimodal data exploration, or new interfaces for musical expression. However, trust in auditory displays is undermined by the non-proportional relationship between synthesizer parameters and perceived sound. In the past decade, AI research into representation learning, creative synthesizer control, and domain transfer applications have opened up new possibilities for modeling human sound perception and for mapping between image and audio data. In this talk, I present how I used AI algorithms during my PhD research to model the perceptual space of sounds produced by a synthesizer, and to project relationships of images into the domain of sound—all in the quest to make image sonification an accessible and reliable instrument for scientific research and sonic design.
Access
MishMash MeetUps are short, informal meetings in the consortium where both early career and established researchers present ongoing projects. The events are open for everyone, but, for security reasons, Zoom links are only provided to people that are affiliated with a MishMash Work Package. If not, please ask for access.
Subscribe to all events (webcal)
Download calendar (ICS)
If the webcal link doesn't open in your calendar (e.g. Outlook), use "Download calendar (ICS)" and open the file in your calendar app.