We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.AS

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Audio and Speech Processing

Title: S2VC: A Framework for Any-to-Any Voice Conversion with Self-Supervised Pretrained Representations

Abstract: Any-to-any voice conversion (VC) aims to convert the timbre of utterances from and to any speakers seen or unseen during training. Various any-to-any VC approaches have been proposed like AUTOVC, AdaINVC, and FragmentVC. AUTOVC, and AdaINVC utilize source and target encoders to disentangle the content and speaker information of the features. FragmentVC utilizes two encoders to encode source and target information and adopts cross attention to align the source and target features with similar phonetic content. Moreover, pre-trained features are adopted. AUTOVC used dvector to extract speaker information, and self-supervised learning (SSL) features like wav2vec 2.0 is used in FragmentVC to extract the phonetic content information. Different from previous works, we proposed S2VC that utilizes Self-Supervised features as both source and target features for VC model. Supervised phoneme posteriororgram (PPG), which is believed to be speaker-independent and widely used in VC to extract content information, is chosen as a strong baseline for SSL features. The objective evaluation and subjective evaluation both show models taking SSL feature CPC as both source and target features outperforms that taking PPG as source feature, suggesting that SSL features have great potential in improving VC.
Comments: Accepted by INTERSPEECH 2021
Subjects: Audio and Speech Processing (eess.AS); Sound (cs.SD)
Cite as: arXiv:2104.02901 [eess.AS]
  (or arXiv:2104.02901v2 [eess.AS] for this version)

Submission history

From: Cheng-Hao Lin [view email]
[v1] Wed, 7 Apr 2021 03:55:53 GMT (224kb,D)
[v2] Mon, 14 Jun 2021 11:10:06 GMT (224kb,D)

Link back to: arXiv, form interface, contact.