FastSpeech2: direct target training with duration/pitch/energy adaptation

What changed

Replaced distilled acoustic targets with ground-truth speech features and added explicit duration, pitch and energy conditioning/prediction, including a waveform-generation extension.

Scientific scope

Replaced distilled acoustic targets with ground-truth speech features and added explicit duration, pitch and energy conditioning/prediction, including a waveform-generation extension. Inherits FastSpeech length regulation and variance modeling. Delta is direct-target/variance-adaptor design, not merely a faster version. External extraction/training costs and error remain; neither2nor2s receives a separate checkpoint-count entry.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Inherits FastSpeech length regulation and variance modeling. Delta is direct-target/variance-adaptor design, not merely a faster version. External extraction/training costs and error remain; neither2nor2s receives a separate checkpoint-count entry.

Contributors

• Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.

• FastSpeech 2: Fast and High-Quality End-to-End Text-to-Speech

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction