Transferable Positive/Negative Speech Emotion Recognition via Class-wise Adversarial Domain Adaptation

2018/10/30 by Hao Zhou, Ke Chen, Zhou, Hao +1
Computer Science · Mathematics · Psychology · #Emotion and Mood Recognition #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Speech Recognition and Synthesis #Speech and Audio Processing #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.1810.12782

5 pages, 3 figures, accepted to ICASSP 2019

openalex publication_date 2018/10/30 · openalex created_date 2018/11/09 · arxiv created 2019/02/14 · arxiv updated 2019/02/15 · openalex updated_date 2026/07/28

Abstract

Speech emotion recognition plays an important role in building more intelligent and human-like agents. Due to the difficulty of collecting speech emotional data, an increasingly popular solution is leveraging a related and rich source corpus to help address the target corpus. However, domain shift between the corpora poses a serious challenge, making domain shift adaptation difficult to function even on the recognition of positive/negative emotions. In this work, we propose class-wise adversarial domain adaptation to address this challenge by reducing the shift for all classes between different corpora. Experiments on the well-known corpora EMODB and Aibo demonstrate that our method is effective even when only a very limited number of target labeled examples are provided.

Citations

• A Survey on Transfer Learning

Related

• Acted vs. Improvised: Domain Adaptation for Elicitation Approaches in Audio-Visual Emotion Recognition
• Emotion-Aware Contrastive Adaptation Network for Source-Free Cross-Corpus Speech Emotion Recognition
• Integrating Contrastive Learning into a Multitask Transformer Model for Effective Domain Adaptation
• Transfer Learning for Improving Speech Emotion Classification Accuracy
• Self Supervised Adversarial Domain Adaptation for Cross-Corpus and Cross-Language Speech Emotion Recognition