Multisource Transfer Learning for Cross-Subject EEG Emotion Recognition

By Jinpeng Li, Shuang Qiu, Yuanyuan Shen, Chenglin Liu, Huiguang He. IEEE Transactions on Cybernetics, 2020

313 citations recorded by the source corpus.

Electroencephalogram (EEG) has been widely used in emotion recognition due to its high temporal resolution and reliability. Since the individual differences of EEG are large, the emotion recognition models could not be shared across persons, and we need to collect new labeled data to train personal models for new users. In some applications, we hope to acquire models for new persons as fast as possible, and reduce the demand for the labeled data amount. To achieve this goal, we propose a multisource transfer learning method, where existing persons are sources, and the new person is the target. The target data are divided into calibration sessions for training and subsequent sessions for test. The first stage of the method is source selection aimed at locating appropriate sources. The second is style transfer mapping, which reduces the EEG differences between the target and each source. We use few labeled data in the calibration sessions to conduct source selection and style transfer. Finally, we integrate the source models to recognize emotions in the subsequent sessions. The experimental results show that the three-category classification accuracy on benchmark SEED improves by 12.72% comparing with the nontransfer method. Our method facilitates the fast deployment of emotion recognition models by reducing the reliance on the labeled data amount, which has practical significance especially in fast-deployment scenarios.

Subfields: Cognitive Neuroscience, Experimental and Cognitive Psychology, Signal Processing

View Paper: https://www.semanticscholar.org/paper/53d0b15342c0bdb908f5da3c4d1afff72c2a4020

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

1. Wei-Long Zheng
2. Wei Liu
3. Dongrui Wu
4. Jielin Qiu
5. Bao-Liang Lu
6. Dehua Liang
7. Elnaz Lashgari
8. Uri Maoz

In Experimental and Cognitive Psychology

1. Wei-Long Zheng
2. Wei Liu
3. Jielin Qiu
4. Dongrui Wu
5. Bao-Liang Lu
6. Dehua Liang
7. Elnaz Lashgari
8. Uri Maoz

In Signal Processing

1. Yudong Zhang
2. Lingfei Deng
3. Jian Wang
4. Dongrui Wu
5. Bao-Liang Lu
6. Wen Zhang
7. Wei-Long Zheng

Citations in the corpus

• Comparing Recognition Performance and Robustness of Multimodal Deep Learning Models for Multimodal Emotion Recognition (2021, methodology): “We select the output dimension from the set [5, 10, 15, 20, 25, 30] and the EEG weight α1 in [0, 0.”
• A Review of Deep Learning on Medical Image Analysis (2020, background): “Cheng, et al. [38] mentioned that transfer learning for early diagnosis of Alzheimer ’ s disease usually adopted all data and annotation from source domains without discriminability of part of [43, 44] would still have a broad prospect on brain medical image processing realm.”
• Transfer Learning for EEG-Based Brain–Computer Interfaces: A Review of Progress Made Since 2016 (2020, unknown): “[92] considered cross-subject EEG emotion classification for both supervised (the target subject has some labelled samples) and semi-supervised (the target subject has some labelled samples and unlabelled samples) scenarios.”
• Overcoming Negative Transfer: A Survey (2020, background): “…[30], and Wasserstein distance [65]; (ii) Selecting discriminative and informative instances, such as removing support vectors gained by SVM [36] and active learning strategy [37]; (iii) Extracting latent space of features form source and target domains to transfer common knowldege, such as…”
• Data Augmentation for Deep-Learning-Based Electroencephalography. (2020, methodology): “To achieve this goal, transfer learning methods have been shown to be helpful [25].”