Publikation
Emo-StarGAN: A Semi-Supervised Any-to-Many Non-Parallel Emotion-Preserving Voice Conversion
Suhita Ghosh; Arnab Das; Yamini Sinha; Ingo Siegert; Tim Polzehl; Sebastian Stober
In: Proc. INTERSPEECH 2023. Conference in the Annual Series of Interspeech Events (INTERSPEECH-2023), Pages 2093-2097, ISCA-speech, 2023.
Zusammenfassung
Speech anonymisation prevents misuse of spoken data by removing any personal identifier while preserving at least linguistic content. However, emotion preservation is crucial for natural human-computer interaction. The well-known voice conversion technique StarGANv2-VC achieves anonymisation but fails to preserve emotion. This work presents an any-to-many semi-supervised StarGANv2-VC variant trained on partially emotion-labelled non-parallel data. We propose emotion-aware losses computed on the emotion embeddings and acoustic features correlated to emotion. Additionally, we use an emotion classifier to provide direct emotion supervision. Objective and subjective evaluations show that the proposed approach significantly improves emotion preservation over the vanilla StarGANv2-VC. This considerable improvement is seen over diverse datasets, emotions, target speakers, and inter-group conversions without compromising intelligibility and anonymisation.
Projekte
- Emonymous - Emonymous - Sprecheranonymisierung für die Privatsphäre und Erschießung von Sprachdaten
- Medinym - KI-basierte Anonymisierung personenbezogener Patientendaten in klinischen Text- und Sprachdatenbeständen