Fully-hierarchical Fine-grained Prosody Modeling For Interpretable Speech Synthesis

Abstract

This paper proposes a hierarchical, fine-grained and interpretable latent variable model for prosody based on the Tacotron 2 text-to-speech model. It achieves multi-resolution modeling of prosody by conditioning finer level representations on coarser level ones. Additionally, it imposes hierarchical conditioning across all latent dimensions using a conditional variational auto-encoder (VAE) with an auto-regressive structure. Evaluation of reconstruction performance illustrates that the new structure does not degrade the model while allowing better interpretability. Interpretations of prosody attributes are provided together with the comparison between word-level and phone-level prosody representations. Moreover, both qualitative and quantitative evaluations are used to demonstrate the improvement in the disentanglement of the latent dimensions.

Stats

• citations58
• S2 citations—
• github stars0
• HF likes0
• heat score13.28
• arxiv keysun2020fully

Related papers

• Hierarchical Prosody Modeling For Non-autoregressive Speech Synthesis (2020)10.07
• Hierarchical Multi-grained Generative Model For Expressive Speech Synthesis (2020)8.60
• Hierarchical Prosody Modeling And Control In Non-autoregressive Parallel Neural TTS (2021)8.35
• Controllable Neural Text-to-speech Synthesis Using Intuitive Prosodic Features (2020)11.76
• Hierarchical Generative Modeling For Controllable Speech Synthesis (2018)0.00
• Unsupervised Quantized Prosody Representation For Controllable Speech Synthesis (2022)4.52
• Towards End-to-end Prosody Transfer For Expressive Speech Synthesis With Tacotron (2018)0.00
• Towards Expressive Zero-shot Speech Synthesis With Hierarchical Prosody Modeling (2024)4.52