Unsupervised acoustic unit discovery for speech synthesis using discrete latent-variable neural networks

no code implementations • 16 Apr 2019 • Ryan Eloff, André Nortje, Benjamin van Niekerk, Avashna Govender, Leanne Nortje, Arnu Pretorius, Elan van Biljon, Ewald van der Westhuizen, Lisa van Staden, Herman Kamper

For our submission to the ZeroSpeech 2019 challenge, we apply discrete latent-variable neural networks to unlabelled speech and use the discovered units for speech synthesis.

Acoustic Unit Discovery Speech Synthesis

A comparison of self-supervised speech representations as input features for unsupervised acoustic word embeddings

no code implementations • 14 Dec 2020 • Lisa van Staden, Herman Kamper

We compare frame-level features from contrastive predictive coding (CPC), autoregressive predictive coding and a CAE to conventional MFCCs.

Representation Learning Word Embeddings