Spatial-Filter-Bank-Based Neural Method for Multichannel Speech Enhancement
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915223235985408 |
|---|---|
| author | Zheng, Tianqin Jin, Jilu Pei, Hanchen Huang, Gongping Chen, Jingdong Benesty, Jacob |
| author_facet | Zheng, Tianqin Jin, Jilu Pei, Hanchen Huang, Gongping Chen, Jingdong Benesty, Jacob |
| contents | The performance of deep learning-based multi-channel speech enhancement methods often deteriorates when the geometric parameters of the microphone array change. Traditional approaches to mitigate this issue typically involve training on multiple microphone arrays, which can be costly. To address this challenge, we focus on uniform circular arrays and propose the use of a spatial filter bank to extract features that are approximately invariant to geometric parameters. These features are then processed by a two-stage conformer-based model (TSCBM) to enhance speech quality. Experimental results demonstrate that our proposed method can be trained on a fixed microphone array while maintaining effective performance across uniform circular arrays with unseen geometric configurations during applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_01392 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Spatial-Filter-Bank-Based Neural Method for Multichannel Speech Enhancement Zheng, Tianqin Jin, Jilu Pei, Hanchen Huang, Gongping Chen, Jingdong Benesty, Jacob Audio and Speech Processing The performance of deep learning-based multi-channel speech enhancement methods often deteriorates when the geometric parameters of the microphone array change. Traditional approaches to mitigate this issue typically involve training on multiple microphone arrays, which can be costly. To address this challenge, we focus on uniform circular arrays and propose the use of a spatial filter bank to extract features that are approximately invariant to geometric parameters. These features are then processed by a two-stage conformer-based model (TSCBM) to enhance speech quality. Experimental results demonstrate that our proposed method can be trained on a fixed microphone array while maintaining effective performance across uniform circular arrays with unseen geometric configurations during applications. |
| title | Spatial-Filter-Bank-Based Neural Method for Multichannel Speech Enhancement |
| topic | Audio and Speech Processing |
| url | https://arxiv.org/abs/2504.01392 |