EEG-based cognitive workload recognition using deep learning techniques

• Main Author: Zhang, Chunlin
• Other Authors: Wang Lipo
• Format: Thesis-Master by Coursework
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/150336
• Institution: Nanyang Technological University

Cognitive workload is an important factor in completing complex cognitive tasks. Cognitive resources are limited and the amount of cognitive resource that devoted to a particular task can seriously affect the performance of complex cognitive tasks. Nowadays, how to effectively detect and evaluate cognitive workload in order to maintain a good working and learning state has been an important research topic. This dissertation mainly studied the existing EEG based cognitive workload recognition algorithms and EEG-based cognitive workload recognition using deep learning techniques among 48 college students. The main contributions include the following aspects: 1.This dissertation reviewed and compared the common cognitive workload recognition deep learning techniques in EEG Analysis and the shortcomings of relevant researches. 2. In the dissertation, we proposed an end-to-end convolutional neural network model based on EEG signals to detect and assess college students’ cognitive workload under high and low workload tasks. 3. This convolutional neural network model get a good accuracy, which is 84.1%. It was a useful exploration on EEG-based cognitive workload recognition. Till now, the present relevant researches are mostly focused on specific types of people, such as vehicle driving and flight research. However, with the popularity of wearable devices, EEG signals will be used in more application scenarios. Therefore, further exploration in deep learning technology applied to EEG signals is very needed.