Investigating training objective for flow matching-based speech enhancement
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866908705012842496 |
|---|---|
| author | Yang, Liusha Ge, Ziru Zhang, Gui Zhang, Junan Wu, Zhizheng |
| author_facet | Yang, Liusha Ge, Ziru Zhang, Gui Zhang, Junan Wu, Zhizheng |
| contents | Speech enhancement(SE) aims to recover clean speech from noisy recordings. Although generative approaches such as score matching and Schrodinger bridge have shown strong effectiveness, they are often computationally expensive. Flow matching offers a more efficient alternative by directly learning a velocity field that maps noise to data. In this work, we present a systematic study of flow matching for SE under three training objectives: velocity prediction, $x_1$ prediction, and preconditioned $x_1$ prediction. We analyze their impact on training dynamics and overall performance. Moreover, by introducing perceptual(PESQ) and signal-based(SI-SDR) objectives, we further enhance convergence efficiency and speech quality, yielding substantial improvements across evaluation metrics. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_10382 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Investigating training objective for flow matching-based speech enhancement Yang, Liusha Ge, Ziru Zhang, Gui Zhang, Junan Wu, Zhizheng Sound Speech enhancement(SE) aims to recover clean speech from noisy recordings. Although generative approaches such as score matching and Schrodinger bridge have shown strong effectiveness, they are often computationally expensive. Flow matching offers a more efficient alternative by directly learning a velocity field that maps noise to data. In this work, we present a systematic study of flow matching for SE under three training objectives: velocity prediction, $x_1$ prediction, and preconditioned $x_1$ prediction. We analyze their impact on training dynamics and overall performance. Moreover, by introducing perceptual(PESQ) and signal-based(SI-SDR) objectives, we further enhance convergence efficiency and speech quality, yielding substantial improvements across evaluation metrics. |
| title | Investigating training objective for flow matching-based speech enhancement |
| topic | Sound |
| url | https://arxiv.org/abs/2512.10382 |