Exploring Self-Attention Mechanisms for Speech Separation
This work addresses speech separation and enhancement for audio processing applications, presenting incremental improvements by extending an existing model to new datasets and tasks.
The paper tackles speech separation by extending the SepFormer Transformer model to noisy and noisy-reverberant datasets like LibriMix, WHAM!, and WHAMR!, and applies it to speech enhancement tasks such as denoising and dereverberation. It introduces efficient self-attention mechanisms like Linformers, Lonformers, and ReFormers, showing that Reformer-based attention outperforms Conv-TasNet on WSJ0-2Mix with faster inference and comparable memory usage.
Transformers have enabled impressive improvements in deep learning. They often outperform recurrent and convolutional models in many tasks while taking advantage of parallel processing. Recently, we proposed the SepFormer, which obtains state-of-the-art performance in speech separation with the WSJ0-2/3 Mix datasets. This paper studies in-depth Transformers for speech separation. In particular, we extend our previous findings on the SepFormer by providing results on more challenging noisy and noisy-reverberant datasets, such as LibriMix, WHAM!, and WHAMR!. Moreover, we extend our model to perform speech enhancement and provide experimental evidence on denoising and dereverberation tasks. Finally, we investigate, for the first time in speech separation, the use of efficient self-attention mechanisms such as Linformers, Lonformers, and ReFormers. We found that they reduce memory requirements significantly. For example, we show that the Reformer-based attention outperforms the popular Conv-TasNet model on the WSJ0-2Mix dataset while being faster at inference and comparable in terms of memory consumption.