Fastspeech: Fast, Robust And Controllable Text To Speech

Abstract

Neural network based end-to-end text to speech (TTS) has significantly improved the quality of synthesized speech. Prominent methods (e.g., Tacotron 2) usually first generate mel-spectrogram from text, and then synthesize speech from the mel-spectrogram using vocoder such as WaveNet. Compared with traditional concatenative and statistical parametric approaches, neural network based end-to-end models suffer from slow inference speed, and the synthesized speech is usually not robust (i.e., some words are skipped or repeated) and lack of controllability (voice speed or prosody control). In this work, we propose a novel feed-forward network based on Transformer to generate mel-spectrogram in parallel for TTS. Specifically, we extract attention alignments from an encoder-decoder based teacher model for phoneme duration prediction, which is used by a length regulator to expand the source phoneme sequence to match the length of the target mel-spectrogram sequence for parallel mel-spectrogram

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyren2019fastspeech

Related papers

• Neural Speech Synthesis With Transformer Network (2018)19.95
• Fastspeech 2: Fast And High-quality End-to-end Text To Speech (2020)0.00
• S-transformer: Segment-transformer For Robust Neural Speech Synthesis (2020)0.00
• Efficienttts: An Efficient And High-quality Text-to-speech Architecture (2020)0.00
• Aligntts: Efficient Feed-forward Text-to-speech System Without Explicit Alignment (2020)11.76
• FPETS : Fully Parallel End-to-end Text-to-speech System (2018)4.52
• JETS: Jointly Training Fastspeech2 And Hifi-gan For End To End Text To Speech (2022)12.10
• STYLER: Style Factor Modeling With Rapidity And Robustness Via Speech Decomposition For Expressive And Controllable Neural Text To Speech (2021)9.23