Inferring the Climate in Classrooms from Audio and Video Recordings: A Machine Learning Approach

• Published in: 2018 IEEE International Conference on Teaching, Assessment, and Learning for Engineering (TALE) pp. 983 - 988
• Main Authors: James, Anusha, Kashyap, Mohan, Chua, Yi Han Victoria, Maszczyk, Tomasz, Nunez, Ana Moreno, Bull, Rebecca, Dauwels, Justin
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2018
• Subjects: audio-video analytics; Australia; classroom climate prediction; Conferences; Education; educational research; Hafnium; machine learning; social behavior
• ISSN: 2470-6698
• DOI: 10.1109/TALE.2018.8615327

The classroom climate is shaped by a combination of teacher practices and peer relationships. The Classroom Assessment Scoring System (CLASS) has been designed to observe and code classroom interactions between students and teachers in order to provide formative feedback on teaching practices and improve teacher instruction. But the turnover time for training, observing and coding makes it hard to generate instant feedback. Since there are few automated assessment tools designed to measure the classroom climate, we propose a novel system for automatic assessment of classroom climate, based on speech, behavioral cues and video features by applying machine learning techniques. This paper elaborates on the design and validation of an audio-video analytics platform for predicting classroom climate. Employing machine learning classifiers instead of subjective measures can ease and expedite the coding. We presume our system can empower education systems to continuously review and improve teaching strategies thus promoting smart classroom in the future.