Pré-Publication, Document De Travail Année : 2025

PITCHFLOWER: A FLOW-BASED NEURAL AUDIO CODEC WITH PITCH CONTROLLABILITY

Résumé

We present PitchFlower, a flow-based neural audio codec with explicit pitch controllability. Our approach enforces disentanglement through a simple perturbation: during training, F0 contours are flattened and randomly shifted, while the true F0 is provided as conditioning. A vector-quantization bottleneck prevents pitch recovery, and a flow-based decoder generates high quality audio. Experiments show that PitchFlower achieves more accurate pitch control than WORLD at much higher audio quality, and outperforms SiFi-GAN in controllability while maintaining comparable quality. Beyond pitch, this framework provides a simple and extensible path toward disentangling other speech attributes. We release the code at https://github.com/diegotg2000/PitchFlower/

Fichier principal
Vignette du fichier
pitchflower_hal.pdf (556.84 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05409018 , version 1 (10-12-2025)

Licence

Identifiants

  • HAL Id : hal-05409018 , version 1

Citer

Diego Torres, Axel Roebel, Nicolas Obin. PITCHFLOWER: A FLOW-BASED NEURAL AUDIO CODEC WITH PITCH CONTROLLABILITY. 2025. ⟨hal-05409018⟩
474 Consultations
173 Téléchargements

Partager

  • More