SEGAN: Speech Enhancement Generative Adversarial Network
This work addresses speech enhancement for noisy audio signals, offering a novel approach that is incremental by applying GANs to this domain.
The authors tackled speech enhancement by proposing SEGAN, a generative adversarial network that operates at the waveform level and is trained end-to-end, achieving viability and effectiveness as confirmed by objective and subjective evaluations on an independent test set with two speakers and 20 noise conditions.
Current speech enhancement techniques operate on the spectral domain and/or exploit some higher-level feature. The majority of them tackle a limited number of noise conditions and rely on first-order statistics. To circumvent these issues, deep networks are being increasingly used, thanks to their ability to learn complex functions from large example sets. In this work, we propose the use of generative adversarial networks for speech enhancement. In contrast to current techniques, we operate at the waveform level, training the model end-to-end, and incorporate 28 speakers and 40 different noise conditions into the same model, such that model parameters are shared across them. We evaluate the proposed model using an independent, unseen test set with two speakers and 20 alternative noise conditions. The enhanced samples confirm the viability of the proposed model, and both objective and subjective evaluations confirm the effectiveness of it. With that, we open the exploration of generative architectures for speech enhancement, which may progressively incorporate further speech-centric design choices to improve their performance.