mlsanityData to 7 Oct 2026

CycleFlow: Purify Information Factors by Cycle Loss

arXiv 2110.09928 ↗First posted 18 Oct 2021eess.AS cs.LG cs.SDcited by 0

Not on the map. Why?

Abstract (arXiv)

SpeechFlow is a powerful factorization model based on information bottleneck (IB), and its effectiveness has been reported by several studies. A potential problem of SpeechFlow, however, is that if the IB channels are not well designed, the resultant factors cannot be well disentangled. In this study, we propose a CycleFlow model that combines random factor substitution and cycle loss to solve this problem. Experiments on voice conversion tasks demonstrate that this simple technique can effectively reduce mutual information among individual factors, and produce clearly better conversion than the IB-based SpeechFlow. CycleFlow can also be used as a powerful tool for speech editing. We demonstrate this usage by an emotion perception experiment.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 33 of its 59 references that are in the map. First posted on arXiv 18 Oct 2021.

Builds on

33 of 59 references.

References · 33

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2017Neural Discrete Representation Learningcited by 4,650
  3. 2016WaveNet: A Generative Model for Raw Audiocited by 1,710
  4. 2018Group Normalizationcited by 1,514
  5. 2017CyCADA: Cycle-Consistent Adversarial Domain Adaptationcited by 1,079
  6. 2018Style Tokens: Unsupervised Style Modeling, Control and Transfer in End-to-End Speech Synthesiscited by 257
  7. 2019AUTOVC: Zero-Shot Voice Style Transfer with Only Autoencoder Losscited by 189
  8. 2019MOSNet: Deep Learning based Objective Assessment for Voice Conversioncited by 126
  9. 2018Translating and Segmenting Multimodal Medical Volumes with Cycle- and Shape-Consistency Generative Adversarial Networkcited by 97
  10. 2018Hierarchical Generative Modeling for Controllable Speech Synthesiscited by 93
  11. 2017Unsupervised Learning of Disentangled and Interpretable Representations from Sequential Datacited by 92
  12. 2020Unsupervised Speech Decomposition via Triple Information Bottleneckcited by 87
21 more
  1. 2019One-shot Voice Conversion by Separating Speaker and Content Representations with Instance Normalizationcited by 82
  2. 2018StarGAN-VC: Non-parallel many-to-many voice conversion with star generative adversarial networkscited by 78
  3. 2021VQMIVC: Vector Quantization and Mutual Information-Based Unsupervised Speech Representation Disentanglement for One-shot Voice Conversioncited by 63
  4. 2016Voice Conversion from Non-parallel Corpora Using Variational Auto-encodercited by 45
  5. 2017Voice Conversion from Unaligned Corpora using Variational Autoencoding Wasserstein Generative Adversarial Networkscited by 38
  6. 2020VQVC+: One-Shot Voice Conversion by Vector Quantization and U-Net architecturecited by 34
  7. 2021Improving Zero-shot Voice Style Transfer via Disentangled Representation Learningcited by 32
  8. 2017Parallel-Data-Free Voice Conversion Using Cycle-Consistent Adversarial Networkscited by 30
  9. 2018Multi-target Voice Conversion without Parallel Data by Adversarially Learning Disentangled Audio Representationscited by 25
  10. 2020Any-to-One Sequence-to-Sequence Voice Conversion using Self-Supervised Discrete Speech Representationscited by 18
  11. 2021S2VC: A Framework for Any-to-Any Voice Conversion with Self-Supervised Pretrained Representationscited by 17
  12. 2019Non-Parallel Voice Conversion with Cyclic Variational Autoencodercited by 15
  13. 2021Adversarially learning disentangled speech representations for robust multi-factor voice conversioncited by 14
  14. 2019Sequence to Sequence Neural Speech Synthesis with Prosody Modification Capabilitiescited by 11
  15. 2018ACVAE-VC: Non-parallel many-to-many voice conversion with auxiliary classifier variational autoencodercited by 9
  16. 2020Many-to-Many Voice Conversion using Conditional Cycle-Consistent Adversarial Networkscited by 9
  17. 2021Global Rhythm Style Transfer Without Text Transcriptionscited by 6
  18. 2017Deep Factorization for Speech Signalcited by 4
  19. 2019Adversarially Trained Autoencoders for Parallel-Data-Free Voice Conversioncited by 3
  20. 2020An empirical analysis of information encoded in disentangled neural speaker representationscited by 2
  21. 2020Disentangled speaker and nuisance attribute embedding for robust speaker verificationcited by 2

Built on by

Not cited yet.