Deep Cross-Corpus Speech Emotion Recognition: Recent Advances and Perspectives
Глубокое распознавание эмоций в речи между корпусами: современные достижения и перспективы
2021-11-29
SCID: 54.1/hsb7b938
Discuss with AI
cross-corpus speech emotion recognitioncross-language speech emotion recognitiondeep learningdomain adaptationhuman-computer interaction
Figures from the paper
Abstract (AI)
Automatic speech emotion recognition (SER) is a challenging component of human-computer interaction (HCI). Existing literatures mainly focus on evaluating the SER performance by means of training and testing on a single corpus with a single language setting. However, in many practical applications, there are great differences between the training corpus and testing corpus. Due to the diversity of different speech emotional corpus or languages, most previous SER methods do not perform well when applied in real-world cross-corpus or cross-language scenarios. Inspired by the powerful feature learning ability of recently-emerged deep learning techniques, various advanced deep learning models have increasingly been adopted for cross-corpus SER. This paper aims to provide an up-to-date and comprehensive survey of cross-corpus SER, especially for various deep learning techniques associated with supervised, unsupervised and semi-supervised learning in this area. In addition, this paper also highlights different challenges and opportunities on cross-corpus SER tasks, and points out its future trends.
Key Findings
1
Cross-corpus and cross-language speech emotion recognition is substantially more difficult than conventional single-corpus, single-language evaluation because training and testing distributions differ.
2
Deep learning techniques are increasingly adopted for cross-corpus speech emotion recognition because of their strong feature-learning capabilities.
3
Many existing speech emotion recognition methods perform poorly in real-world cross-corpus and cross-language scenarios due to diversity across speech emotion corpora and languages.
4
The paper identifies major challenges, opportunities, and future research directions for improving cross-corpus speech emotion recognition.
5
The survey comprehensively reviews supervised, unsupervised, and semi-supervised deep learning approaches for cross-corpus speech emotion recognition.
Research Object
cross-corpus and cross-language automatic speech emotion recognition systems
Research Subject
deep-learning-based methods, performance, challenges, and future directions for recognizing speech emotions across differing training and testing corpora or languages
Publication Details
Publication Date
2021-11-29
Journal
Publisher
ISSN
Cited by
48
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai4
ImageNet classification with deep convolutional neural networks2017
Reducing the Dimensionality of Data with Neural Networks2006
A survey on semi-supervised learning2019
The Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS): A dynamic, multimodal set of facial and vocal expressions in North American English2018