Word Segmentation on Discovered Phone Units with Dynamic Programming and Self-Supervised Scoring
This work addresses unsupervised speech processing for low-resource languages, though it is incremental as it revisits and refines older segmentation approaches.
The paper tackles unsupervised speech word segmentation by chaining bottom-up phone-like unit discovery with symbolic word segmentation using dynamic programming and self-supervised scoring, achieving comparable results to state-of-the-art joint models on English and outperforming previous systems on French, Mandarin, German, and Wolof ZeroSpeech benchmarks.
Recent work on unsupervised speech segmentation has used self-supervised models with phone and word segmentation modules that are trained jointly. This paper instead revisits an older approach to word segmentation: bottom-up phone-like unit discovery is performed first, and symbolic word segmentation is then performed on top of the discovered units (without influencing the lower level). To do this, I propose a new unit discovery model, a new symbolic word segmentation model, and then chain the two models to segment speech. Both models use dynamic programming to minimize segment costs from a self-supervised network with an additional duration penalty that encourages longer units. Concretely, for acoustic unit discovery, duration-penalized dynamic programming (DPDP) is used with a contrastive predictive coding model as the scoring network. For word segmentation, DPDP is applied with an autoencoding recurrent neural as the scoring network. The two models are chained in order to segment speech. This approach gives comparable word segmentation results to state-of-the-art joint self-supervised segmentation models on an English benchmark. On French, Mandarin, German and Wolof data, it outperforms previous systems on the ZeroSpeech benchmarks. Analysis shows that the chained DPDP system segments shorter filler words well, but longer words might require some external top-down signal.