Kanade: A Simple Disentangled Tokenizer for Spoken Language Modeling
This work addresses the need for effective speech tokenization in spoken language modeling, enabling better language models by extracting phonetics and prosody while suppressing speaker identity.
Kanade is a single-layer disentangled speech tokenizer that separates linguistic content from speaker identity without auxiliary methods, achieving state-of-the-art speaker disentanglement and lexical availability while maintaining high reconstruction quality.
A good language model starts with a good tokenizer. Tokenization is especially important for speech modeling, which must handle continuous signals that mix linguistic and non-linguistic information. A speech tokenizer should extract phonetics and prosody, suppress linguistically irrelevant information like speaker identity, and enable high-quality synthesis. We present Kanade, a single-layer disentangled speech tokenizer that realizes this ideal. Kanade separates out acoustic constants to create a single stream of tokens that captures rich phonetics and prosody. It does so without the need for auxiliary methods that existing disentangled codecs often rely on. Experiments show that Kanade achieves state-of-the-art speaker disentanglement and lexical availability, while maintaining excellent reconstruction quality.