AffecTeach

At the core of the project, an AI-assisted simulator was developed to train teachers. This was for trainee teachers to practice with virtual students that can react according to their emotional and cognitive stances, and in return the trainees get objective performance feedback.

This project composed of the following major objectives: Preparing student talk schemes based on the analysis of classroom recordings; Automatically recognising teacher's emotional and cognitive responses via multi-modal data processing; Constructing the simulator and integrating with student talk schemes and automatic emotion recognition modules; Implementing a recommendation and feedback system for teachers' professional development.

This interdisciplinary project was supported by an Institutional Links grant, under the Newton-Katip Celebi partnership. Loughborough University London (Institute for Digital Technologies) partnered with Hacettepe University (Department of Computer Education and Instructional Technology) and SimSoft Ltd from Turkey. The technical objectives of Loughborough University London concentrated around multimodal emotion recognition system design and its integration with the training simulator.

This was a two-year project, which ended in 2020. Notable publications from this project, which feature Loughborough University London based authors include:

S. Caglar-Ozhan, A. Altun, E. Ekmekcioglu, “The Investigation of Prospective Teachers’ Emotions and Emotional Patterns in a Simulated Virtual Classroom Supported with an Affective Recommendation System”, British Journal of Educational Technology, 2021 (under review).

Y. Cimtay, E. Ekmekcioglu, S. Caglar-Ozhan, "Cross-Subject Multimodal Emotion Recognition Based on Hybrid Fusion," IEEE Access, vol. 8, pp. 168865-168878, 2020.

Y. Cimtay, E. Ekmekcioglu, “Investigating the use of pre-trained Convolutional Neural Network on Cross-Subject and Cross-Dataset EEG Emotion Recognition”, Sensors, 20 (7), 2034, 2020.