Abstract
Recent advances in generating synthetic captions based on audio and related metadata allow using the information contained in natural language as input for other audio tasks. In this paper, we propose a novel method to guide a sound event detection system with free-form text. We use machine-generated captions as complementary information to the strong labels for training, and evaluate the systems using different types of textual inputs. In addition, we study a scenario where only part of the training data has strong labels, and the rest of it only has temporally weak labels. Our findings show that synthetic captions improve the performance in both cases compared to the CRNN architecture typically used for sound event detection. On a dataset of 50 highly unbalanced classes, the PSDS-1 score increases from 0.223 to 0.277 when trained with strong labels, and from 0.166 to 0.218 when half of the training data has only weak labels.
| Original language | English |
|---|---|
| Title of host publication | 2025 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA) |
| Publisher | IEEE |
| ISBN (Electronic) | 979-8-3315-3745-6 |
| DOIs | |
| Publication status | Published - 2025 |
| Publication type | A4 Article in conference proceedings |
| Event | IEEE Workshop on Applications of Signal Processing to Audio and Acoustics - Tahoe City, United States Duration: 12 Oct 2025 → 15 Oct 2025 |
Publication series
| Name | IEEE Workshop on Applications of Signal Processing to Audio and Acoustics |
|---|---|
| ISSN (Electronic) | 1947-1629 |
Conference
| Conference | IEEE Workshop on Applications of Signal Processing to Audio and Acoustics |
|---|---|
| Country/Territory | United States |
| City | Tahoe City |
| Period | 12/10/25 → 15/10/25 |
Funding
This work was supported by Academy of Finland grant 332063 "Teaching machines to listen". The authors wish to thank CSC-IT Centre of Science Ltd., Finland, for providing computational resources.
Publication forum classification
- Publication forum level 1
Fingerprint
Dive into the research topics of 'Sound event detection with audio-text models and heterogeneous temporal annotations'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver