A Concentration Perception Method for Learner Learning Process Based on Multi-space Fusion

Abstract

The invention discloses a method for perceiving concentration of learners in the learning process based on multi-space fusion. The method includes: acquiring the perceived temperature image of the students and teachers during classroom learning; preprocessing the perceived temperature image, and selecting a key frame of the perceived temperature image according to a preset frequency; inputting the selected key frame 2D convolutional neural network to obtain the first output feature; obtain the mouse movement track during the online learning process of the students; input the mouse movement track with time series into the 3D convolutional neural network to obtain the second output feature; The first output feature and the second output feature are input into the SVM linear classifier to obtain the concentration classification result. The present invention realizes the use of deep learning to extract features, extracts different features from the two-dimensional convolutional network and the three-dimensional convolutional network, and uses the SVM linear classifier to classify the features, so that better classification effects can be obtained.

Description

technical field [0001] This application relates to the field of educational informatization technology, in particular, to a method for perceiving concentration of learners in the learning process based on multi-space fusion. Background technique [0002] Temperature is a characteristic that everyone has all the time in life, and it is also a characteristic that appears in every part of the human body. At the same time, when there are different fluctuations in the psychological characteristics or physical state of the person, the temperature of different parts of the human body will appear. Variety. Therefore, the temperature of a person's body part will directly reflect the learning status of the student. In the learning process of cyberspace, the movement of the mouse is also an essential human activity, and different mouse movement trajectories also reflect the current learning status, concentration and other information of students. Therefore, capturing these natural re...

AI Technical Summary

Problems solved by technology

However, head posture is an external behavior in the learning process of learners, and it is easy to generate disturbing data due to personal habits, which will affect the concentration detection results

Images