Teaching Assistant: A Classroom Teaching Effectiveness Evaluation System

Classroom teaching is the center stage for training talents. The main goal of classroom teaching is to improve teaching quality and achieve efficient teaching activities. Classroom teaching is a bilateral activity of teachers' teaching and students' learning. It should not only pay attention to teachers but also to the students. However, in the scene of traditional classroom teaching, it is difficult for teachers to monitor every student's learning status and accurately understand the overall learning status of the classroom in real time. So, the teaching effect is difficult to ensure. In order to help teachers carry out teaching more flexibly and effectively, this paper proposes a new classroom teaching effectiveness evaluation system-Teaching Assistant. Firstly, this study employs the MTCNN network to achieve facial detection in the classroom. This network model exhibits certain stability in complex classroom environments and can effectively detect students' facial information. Secondly, based on the mini_Xception architecture, this study aims to recognize the expressions of students', which guarantees recognition accuracy while effectively reducing the model's calculation cost, making it more practical to classroom environments. Finally, this study proposes a student status evaluation method to analyze students' emotional states and confusion levels by calculating the overall learning status statistics of the classroom. The experimental results show that Teaching Assistant based on classroom teaching is beneficial for teachers to know both of the individual student's and the overall classroom's learning status in real-time, and to carry out teaching activities more targetedly.