A TPACK-Aligned Teaching Approach: Image Recognition via Convolutional Neural Networks for Artificial Intelligence Education in Teacher Training

Convolutional Neural Networks (CNNs) are the technology behind many familiar applications such as facial recognition, autonomous driving, medical imaging, and augmented reality. This paper introduces a practice-based approach designed to help ICT and digital education educators teach CNNs effectively by combining both theoretical concepts and hands-on activities. Leveraging established educational frameworks like TPACK for teacher competencies and Bloom’s Taxonomy for learning goals on different cognitive levels to ensure all important aspects about CNNs are covered. Our approach includes engaging demonstrations, practical programming exercises, and critical discussions on real-world applications and issues such as transparency, explainability, and data quality in Neural Network training.

Preprint; Publication only accesible at https://www.learntechlib.org/p/226171/