From Acoustic Scene Analysis to Continual Learning for Audio
About this Topic:
This talk will give a brief overview of selected research topics in Detection and Classification of Acoustic Scenes and Events (DCASE). DCASE is a broad research area focused on the analysis of everyday sound scenes, aiming to extract information about acoustic environments using audio signal processing and machine learning methods. Just as humans interpret the sounds around them to understand their surroundings, a main goal in DCASE is to develop systems that can interpret acoustic environments in a similar way.
The first part of the talk will introduce environmental sound analysis from a general perspective, including our experience in data collection and a short overview of the DCASE Challenge. The second part will present continual learning for audio, including our work on class- and domain-incremental learning for sound and acoustic scene classification, as well as a brief analysis of the first task on incremental learning organized within DCASE 2026 Challenge.
About the Presenter:
Annamaria Mesaros (SM’26) received the Ph.D. in signal processing from Tampere University of Technology in 2012.
She is currently an Associate Professor at Tampere University. Her research focuses on sound event detection in real-world multisource environments and includes over 50 scientific publications on this topic and many open datasets.
Dr. Mesaros is a member of the Audio and Acoustic Signal Processing Technical Committee of IEEE Signal Processing Society and of the EURASIP Acoustic, Speech, and Music Signal Processing Technical Area Committee. She is a coordinator of the Detection and Classification of Acoustic Scenes and Events (DCASE) Challenge, vice-chair of the DCASE Steering Group, an associate faculty of ELLIS Institute Finland, and currently an Academy of Finland Research Fellow for "Teaching Machines to Listen".
Want to learn more about upcoming events & webinars?
Visit the events section of the Signal Processing website to see all upcoming lectures, workshops, webinars, and more.