Deep Learning-Based Assessment of Sports-Assisted Teaching and Learning

The current Internet development situation regarding the analysis of sports teaching information is very necessary and can be a way to improve effectiveness in environment. Aiming at defects strong subjectivity low discrimination accuracy video classification results, this paper proposes an effective method based on deep learning, which effectively evaluate assisted teaching. Specifically, key frame features are obtained by using similarity coefficient extraction algorithm, image established through learning coding model. Thus, ability school rely scheme proposed facilities, physical education curriculum materials, assessment management, so on. results show that for different types videos, overall effect significantly better than other sports-assisted evaluation methods, has improved evaluation.