Saved in:
Bibliographic Details
Main Authors: Chen, Yun, Yang, Lingxiao, Chen, Qi, Lai, Jian-Huang, Xie, Xiaohua
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2312.17508
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914624612335616
author Chen, Yun
Yang, Lingxiao
Chen, Qi
Lai, Jian-Huang
Xie, Xiaohua
author_facet Chen, Yun
Yang, Lingxiao
Chen, Qi
Lai, Jian-Huang
Xie, Xiaohua
contents Emotional Voice Conversion aims to manipulate a speech according to a given emotion while preserving non-emotion components. Existing approaches cannot well express fine-grained emotional attributes. In this paper, we propose an Attention-based Interactive diseNtangling Network (AINN) that leverages instance-wise emotional knowledge for voice conversion. We introduce a two-stage pipeline to effectively train our network: Stage I utilizes inter-speech contrastive learning to model fine-grained emotion and intra-speech disentanglement learning to better separate emotion and content. In Stage II, we propose to regularize the conversion with a multi-view consistency mechanism. This technique helps us transfer fine-grained emotion and maintain speech content. Extensive experiments show that our AINN outperforms state-of-the-arts in both objective and subjective metrics.
format Preprint
id arxiv_https___arxiv_org_abs_2312_17508
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Attention-based Interactive Disentangling Network for Instance-level Emotional Voice Conversion
Chen, Yun
Yang, Lingxiao
Chen, Qi
Lai, Jian-Huang
Xie, Xiaohua
Audio and Speech Processing
Artificial Intelligence
Sound
Emotional Voice Conversion aims to manipulate a speech according to a given emotion while preserving non-emotion components. Existing approaches cannot well express fine-grained emotional attributes. In this paper, we propose an Attention-based Interactive diseNtangling Network (AINN) that leverages instance-wise emotional knowledge for voice conversion. We introduce a two-stage pipeline to effectively train our network: Stage I utilizes inter-speech contrastive learning to model fine-grained emotion and intra-speech disentanglement learning to better separate emotion and content. In Stage II, we propose to regularize the conversion with a multi-view consistency mechanism. This technique helps us transfer fine-grained emotion and maintain speech content. Extensive experiments show that our AINN outperforms state-of-the-arts in both objective and subjective metrics.
title Attention-based Interactive Disentangling Network for Instance-level Emotional Voice Conversion
topic Audio and Speech Processing
Artificial Intelligence
Sound
url https://arxiv.org/abs/2312.17508