Combining Masked Language Modeling and Cross-Modal Contrastive Learning for Prosody-Aware TTS

arXiv:2604.01247v2 Announce Type: replace Abstract: Text encoders in neural text-to-speech (TTS) see only linguistic content, so prosody must be inferred from paired data and synthesised intonation is often flat. Contrastive text-audio pretraining (CLAPSpeech) injects prosodic knowledge by contrasting acoustic realisations of the same phoneme, but it never trains the encoder to tell different phonemes apart, and its interaction with masked language modelling (MLM) is unexplored. We study a curriculum for a speaker-conditioned phoneme-and-subword encoder: MLM, contrastive alignment over batches of different phonemes, and CLAPSpeech-style same-phoneme refinement. On Russian conversational speech, with each variant conditioning an identical Grad-TTS model, MLM followed by mixed-phoneme alignment gives the lowest word error rate (17.6% versus 32.8% for same-phoneme training alone) and the best quality and intonation ratings. Adding same-phoneme refinement improves same-phoneme retrieval yet yields the worst synthesis: embedding-space retrieval does not predict usefulness for generation.