Dynamic Frequency-Adaptive Knowledge Distillation for Speech Enhancement
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866929702206177280 |
|---|---|
| author | Yuan, Xihao Liu, Siqi Chen, Hanting Zhou, Lu Li, Jian Hu, Jie |
| author_facet | Yuan, Xihao Liu, Siqi Chen, Hanting Zhou, Lu Li, Jian Hu, Jie |
| contents | Deep learning-based speech enhancement (SE) models have recently outperformed traditional techniques, yet their deployment on resource-constrained devices remains challenging due to high computational and memory demands. This paper introduces a novel dynamic frequency-adaptive knowledge distillation (DFKD) approach to effectively compress SE models. Our method dynamically assesses the model's output, distinguishing between high and low-frequency components, and adapts the learning objectives to meet the unique requirements of different frequency bands, capitalizing on the SE task's inherent characteristics. To evaluate the DFKD's efficacy, we conducted experiments on three state-of-the-art models: DCCRN, ConTasNet, and DPTNet. The results demonstrate that our method not only significantly enhances the performance of the compressed model (student model) but also surpasses other logit-based knowledge distillation methods specifically for SE tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_04711 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Dynamic Frequency-Adaptive Knowledge Distillation for Speech Enhancement Yuan, Xihao Liu, Siqi Chen, Hanting Zhou, Lu Li, Jian Hu, Jie Sound Audio and Speech Processing Deep learning-based speech enhancement (SE) models have recently outperformed traditional techniques, yet their deployment on resource-constrained devices remains challenging due to high computational and memory demands. This paper introduces a novel dynamic frequency-adaptive knowledge distillation (DFKD) approach to effectively compress SE models. Our method dynamically assesses the model's output, distinguishing between high and low-frequency components, and adapts the learning objectives to meet the unique requirements of different frequency bands, capitalizing on the SE task's inherent characteristics. To evaluate the DFKD's efficacy, we conducted experiments on three state-of-the-art models: DCCRN, ConTasNet, and DPTNet. The results demonstrate that our method not only significantly enhances the performance of the compressed model (student model) but also surpasses other logit-based knowledge distillation methods specifically for SE tasks. |
| title | Dynamic Frequency-Adaptive Knowledge Distillation for Speech Enhancement |
| topic | Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2502.04711 |