Self-supervised Predictive Coding Models Encode Speaker and Phonetic Information in Orthogonal Subspaces
This work addresses the challenge of speaker variability in speech processing for applications like speech recognition, though it is incremental as it builds on existing predictive coding models.
The paper tackled the problem of disentangling speaker and phonetic information in self-supervised speech representations by hypothesizing they are encoded in orthogonal subspaces, and it resulted in a speaker normalization method that outperformed a baseline in phone discrimination tasks and generalized to unseen speakers.
Self-supervised speech representations are known to encode both speaker and phonetic information, but how they are distributed in the high-dimensional space remains largely unexplored. We hypothesize that they are encoded in orthogonal subspaces, a property that lends itself to simple disentanglement. Applying principal component analysis to representations of two predictive coding models, we identify two subspaces that capture speaker and phonetic variances, and confirm that they are nearly orthogonal. Based on this property, we propose a new speaker normalization method which collapses the subspace that encodes speaker information, without requiring transcriptions. Probing experiments show that our method effectively eliminates speaker information and outperforms a previous baseline in phone discrimination tasks. Moreover, the approach generalizes and can be used to remove information of unseen speakers.