Mixup augmentation for generalizable speech separation - Ashish Alex
2021 Intelligent Sensing Winter School
Mixup augmentation for generalizable speech separation
Ashish Alex, Queen Mary University of London
Deep learning has advanced the state of the art of single-channel speech separation. However, separation models may overfit the training data and generalization across datasets is still an open problem in real-world conditions with noise. In our paper we addressed the generalization problem with Mixup as data augmentation approach. Mixup creates new training examples from linear combinations of samples during mini-batch training. We proposed four variations of Mixup and assess the improved generalization of a speech separation model, DPRNN, with cross-corpus evaluation on LibriMix, TIMIT and VCTK datasets. We show that training DPRNN with the proposed Data-only Mixup augmentation variation improves performance on an unseen dataset in noisy conditions when compared to the baseline SpecAugment augmented models, while having comparable performance on the source dataset.
Slides: http://cis.eecs.qmul.ac.uk/2021Winter...
As part of the 2021 Intelligent Sensing Winter School: http://cis.eecs.qmul.ac.uk/school2021...