Hybrid Model of Eye Movement Behavior Recognition for Virtual Workshop

Abstract

Purpose This paper is oriented to the virtual workshop scenario, discusses the creation of a feasible eye movement recognition model and provides new ideas for eye movement recognition and eye movement interaction in the virtual workshop. Method Obtain user’s eye movement data through eye tracker in real time, create blink behavior recognition model and gaze/saccade behavior recognition model based on convolutional neural network (CNN) and bidirectional long short-term memory network (Bi-LSTM). A data set was created by collecting eye movement data of three subjects for comparative experiments to verify the performance of the proposed model. Results The experimental results showed that the recognition accuracy, Kappa coefficient, F1 score and running time of the proposed model had certain advantages over other models. Conclusion The created hybrid model of eye movement recognition for virtual workshops has high reliability and effectiveness, which will create a foundation for future research on building eye movement interaction systems.