A Novel Deep Learning Technique for Higher Education System

This study aims to enhance the interest of university classrooms and learning efficiency students. From deep learning, gesture recognition algorithms (GRA) long short-term memory (LSTM) are adopted build prediction modules. After system process is completed, a brand-new human-computer interaction (HCI) established. The GRA augments operation algorithm pattern matching method created on previous image segmentation. Its enticements two merits estimation as follows: rapid detection joint features convolutional neural network (CNN) classification, thereby improving GRA. rates “take”, “pinch”, “point” 98.6%, 99.5%, 99.4%, respectively. When data volume LSTM model less than 10, accuracy can reach 70%, performance relatively stable. designed HCI better recognize teacher’s intent in teaching, execute commands, optimize interactive teaching between teacher computer. derivative form gamification teaching. While it improves also great significance for general higher education.