Where's That Voice Coming? Continual Learning for Sound Source Localization
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916657524375552 |
|---|---|
| author | Xiao, Yang Das, Rohan Kumar |
| author_facet | Xiao, Yang Das, Rohan Kumar |
| contents | Sound source localization (SSL) is essential for many speech-processing applications. Deep learning models have achieved high performance, but often fail when the training and inference environments differ. Adapting SSL models to dynamic acoustic conditions faces a major challenge: catastrophic forgetting. In this work, we propose an exemplar-free continual learning strategy for SSL (CL-SSL) to address such a forgetting phenomenon. CL-SSL applies task-specific sub-networks to adapt across diverse acoustic environments while retaining previously learned knowledge. It also uses a scaling mechanism to limit parameter growth, ensuring consistent performance across incremental tasks. We evaluated CL-SSL on simulated data with varying microphone distances and real-world data with different noise levels. The results demonstrate CL-SSL's ability to maintain high accuracy with minimal parameter increase, offering an efficient solution for SSL applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_03661 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Where's That Voice Coming? Continual Learning for Sound Source Localization Xiao, Yang Das, Rohan Kumar Audio and Speech Processing Sound Sound source localization (SSL) is essential for many speech-processing applications. Deep learning models have achieved high performance, but often fail when the training and inference environments differ. Adapting SSL models to dynamic acoustic conditions faces a major challenge: catastrophic forgetting. In this work, we propose an exemplar-free continual learning strategy for SSL (CL-SSL) to address such a forgetting phenomenon. CL-SSL applies task-specific sub-networks to adapt across diverse acoustic environments while retaining previously learned knowledge. It also uses a scaling mechanism to limit parameter growth, ensuring consistent performance across incremental tasks. We evaluated CL-SSL on simulated data with varying microphone distances and real-world data with different noise levels. The results demonstrate CL-SSL's ability to maintain high accuracy with minimal parameter increase, offering an efficient solution for SSL applications. |
| title | Where's That Voice Coming? Continual Learning for Sound Source Localization |
| topic | Audio and Speech Processing Sound |
| url | https://arxiv.org/abs/2407.03661 |