Multi-View Attention Transfer for Efficient Speech Enhancement
This work addresses the problem of efficient deployment of speech enhancement models for applications requiring low computational resources, though it is incremental as it builds on existing knowledge distillation techniques.
The paper tackled the challenge of creating fast, low-complexity speech enhancement models without performance loss by proposing multi-view attention transfer (MV-AT), a feature-based distillation method that improved student models on datasets like Valentini and DNS, achieving up to 15.4x fewer parameters and 4.71x fewer FLOPs with similar performance.
Recent deep learning models have achieved high performance in speech enhancement; however, it is still challenging to obtain a fast and low-complexity model without significant performance degradation. Previous knowledge distillation studies on speech enhancement could not solve this problem because their output distillation methods do not fit the speech enhancement task in some aspects. In this study, we propose multi-view attention transfer (MV-AT), a feature-based distillation, to obtain efficient speech enhancement models in the time domain. Based on the multi-view features extraction model, MV-AT transfers multi-view knowledge of the teacher network to the student network without additional parameters. The experimental results show that the proposed method consistently improved the performance of student models of various sizes on the Valentini and deep noise suppression (DNS) datasets. MANNER-S-8.1GF with our proposed method, a lightweight model for efficient deployment, achieved 15.4x and 4.71x fewer parameters and floating-point operations (FLOPs), respectively, compared to the baseline model with similar performance.