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: Utilizing Self-supervised Representations for MOS Prediction

Abstract: Speech quality assessment has been a critical issue in speech processing for decades. Existing automatic evaluations usually require clean references or parallel ground truth data, which is infeasible when the amount of data soars. Subjective tests, on the other hand, do not need any additional clean or parallel data and correlates better to human perception. However, such a test is expensive and time-consuming because crowd work is necessary. It thus becomes highly desired to develop an automatic evaluation approach that correlates well with human perception while not requiring ground truth data. In this paper, we use self-supervised pre-trained models for MOS prediction. We show their representations can distinguish between clean and noisy audios. Then, we fine-tune these pre-trained models followed by simple linear layers in an end-to-end manner. The experiment results showed that our framework outperforms the two previous state-of-the-art models by a significant improvement on Voice Conversion Challenge 2018 and achieves comparable or superior performance on Voice Conversion Challenge 2016. We also conducted an ablation study to further investigate how each module benefits the task. The experiment results are implemented and reproducible with publicly available toolkits.
Comments: In Proceedings of Interspeech 2021. We acknowledge the support of AWS Machine Learning Research Awards program. Source code available at this https URL
Subjects: Audio and Speech Processing (eess.AS); Machine Learning (cs.LG); Sound (cs.SD)
Cite as: arXiv:2104.03017 [eess.AS]
  (or arXiv:2104.03017v3 [eess.AS] for this version)

Submission history

From: Wei-Cheng Tseng [view email]
[v1] Wed, 7 Apr 2021 09:44:36 GMT (515kb,D)
[v2] Wed, 21 Apr 2021 11:31:40 GMT (515kb,D)
[v3] Mon, 20 Sep 2021 14:12:43 GMT (515kb,D)

Link back to: arXiv, form interface, contact.