A Simple HMM with Self-Supervised Representations for Phone Segmentation
This work addresses phone segmentation for speech processing, but it is incremental as it builds on existing methods with a new formulation.
The paper tackled unsupervised phonetic segmentation by showing that peak detection on Mel spectrograms outperforms many self-supervised methods, and proposed a simple hidden Markov model using self-supervised representations and boundary features, achieving consistent improvements over previous approaches.
Despite the recent advance in self-supervised representations, unsupervised phonetic segmentation remains challenging. Most approaches focus on improving phonetic representations with self-supervised learning, with the hope that the improvement can transfer to phonetic segmentation. In this paper, contrary to recent approaches, we show that peak detection on Mel spectrograms is a strong baseline, better than many self-supervised approaches. Based on this finding, we propose a simple hidden Markov model that uses self-supervised representations and features at the boundaries for phone segmentation. Our results demonstrate consistent improvements over previous approaches, with a generalized formulation allowing versatile design adaptations.