Improving Stereo 3D Sound Event Localization and Detection: Perceptual Features, Stereo-specific Data Augmentation, and Distance Normalization
2025
Read Improving Stereo 3D Sound Event Localization and Detection: Perceptual Features, Stereo-specific Data Augmentation, and Distance Normalization (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-00874.txt.
Description
Jun-Wei Yeow and co-authors report their entry to a 2025 challenge on locating and naming sounds in stereo video audio, in this arXiv technical report. The team designed input features shaped by how hearing works, lifting detection, direction finding, and distance guesses at once. They also tailored data augmentation to stereo, swapping channels and masking patches of the sound picture.
A further trick, distance normalization during training, steadied the numbers the model must predict. On the STARSS23 test set the changes stacked into steady gains across every scoring measure of the task. The authors share their code so others can rebuild the system. The work targets everyday video content rather than studio recordings.
Catalog record
| YEAR | 2025 |
|---|---|
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 15 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-00874