Research On Student Evaluation System Aided By Video Recognition

The evaluation of learning status is a very important part of education.In the evaluation,we should focus on the advantages and potentials of each student.It is necessary to pay attention to the degree of interest,concentration and participation of students in the course or activities.But at present,the evaluation of students is still very unsound.For example,most of the evaluation process is to strengthen the volume fraction of memory.In addition,only by relying on teachers can we not track and judge each student's learning process.Video recognition assisted student assessment technology can help educators achieve a full objective evaluation of students' learning state to a certain extent.In the evaluation of students' learning state,interest in learning is one of the important indicators.Interest in learning is mainly manifested in the smiling state of students in class or activities.It is considered that some of the students' smile is related to the content of learning,and some have nothing to do with the content of learning,but the irrelevant smile directly affects the judgment of the students' interest in learning.Therefore,this paper focuses on a video recognition aided student evaluation system,especially the method of evaluating the students' smile effectiveness(interest)in teaching activities.The operation process of the system is as follows:first,collect the students' video in the teaching activities,select the effective frames of the video clip through the face detection,and analyze the 68 feature points of the selected effective frames to obtain the characteristics of the students' learning emotion(the emphasis is the smile intensity in the students' learning process).Support vector machine(SVM)is used to classify students' interest in learning in class.The main work of this article is as follows:(1)Through multiple verification,a video acquisition hardware platform for students' learning state evaluation is built to ensure the integrity of data acquisition and the flexibility of building the collection platform.(2)A classification algorithm of smiling effectiveness(according to the relevance of interest degree)in students' teaching activities based on SVM and smile intensity is proposed to help determine the students' interest in learning content.(3)The function of student evaluation system based on video recognition is completed,and the effectiveness of the algorithm is verified.