A Comparison of Discrete Latent Variable Models for Speech Representation Learning
This work addresses speech representation learning for researchers, but it is incremental as it compares existing methods.
The paper compared vq-vae and vq-wav2vec for speech representation learning, finding that vq-wav2vec, based on future time-step prediction, performed better, achieving a 13.22 error rate on the ZeroSpeech 2019 ABX challenge.
Neural latent variable models enable the discovery of interesting structure in speech audio data. This paper presents a comparison of two different approaches which are broadly based on predicting future time-steps or auto-encoding the input signal. Our study compares the representations learned by vq-vae and vq-wav2vec in terms of sub-word unit discovery and phoneme recognition performance. Results show that future time-step prediction with vq-wav2vec achieves better performance. The best system achieves an error rate of 13.22 on the ZeroSpeech 2019 ABX phoneme discrimination challenge