IBAC Online Symposium 2022 - Frontiers in Bioacoustics

Friday 9th September 2022, 07:00 – 11:00 UTC

The IBAC Online Symposium 2022 was a fully virtual event taking place on Zoom. The symposium featured the latest advances in bioacoustics while facilitating exchange among IBAC members. Videos from the meeting.
Sponsored by:
Wildlife Acoustics logo

Follow us on twitter for updates and news about the event: @2022ibac

Download the poster for the event here, please distribute and share with all you think may be interested in joining: JPG PDF

Access to the online presentations is free for everyone but attendees need to register by completing the form below. IBAC members will enjoy the additional opportunity to interact at the symposium and engage in discussions. If you are not a member, you can join here.
Registration deadline: Monday 5 September 23:59 UTC



We plan an exciting and stimulating event with several invited speakers (all times UTC):

07:00-07:40    Hannah ter Hofstede
07:40-08:20    Naomi Langmore
08:20-08:30    Break
08:30-09:10    Dan Stowell
09:10-09:50    Stephanie King
09:50-10:00    Break
10:00-10:40    Iris Adam
10:40-10:50    IBAC Society news and updates (IBAC Committee)
10:50-10:55    Info on Bioacoustic Stack Exchange (Thejasvi Beleyur)
10:55-11:00    Wrap up and Farewell

Hannah ter HofstedeMultimodal duetting in lebinthine crickets: evolution and function
Hannah ter Hofstede
Associate Professor of Biological Sciences, Dartmouth College, Hanover NH, USA
Singing Female Superb Fairy-wren - Lindley McKayBirdsong: the female perspective
Naomi Langmore

Australian National University

Bioacoustic deep learning: new methods and difficult audio datasets
Dan Stowell Dan Stowell
Associate Professor of AI & Biodiversity, Tilburg University, Naturalis Biodiversity Centre, JADS, The Netherlands

Deep learning is increasingly being used to get meaningful results from animal sound recordings: it can handle detection/classification at large scale, when trained with a well-labelled dataset, using neural networks that are now maturing and easy to use. However, everything is not yet easy. Many bioacoustic studies and datasets come with issues that demand more than a "standard" AI model can give: such as detailed within-species discriminations, high accuracy requirements, unlabelled or rarely-occurring sounds, varying conditions, or dense soundscapes. In this talk Dan will briefly survey the state of the art, then describe recent methodological work to bring the capability of deep learning closer to the realities of bioacoustic sound scenes. He will discuss classification and detection, and also describe recent work on few-shot sound event detection.

Stephanie KingVocal and social complexity in bottlenose dolphins
Stephanie King
Associate Professor in Animal Behaviour, University of Bristol, United Kingdom

SyrinxIris AdamSongbirds need daily song to maintain peak vocal performance
Iris Adam
Assistant Professor, University of Southern Denmark