Learning interpretable concept groups in CNNs

• Main Authors: VARSHNEYA, Saurabh, LEDENT, Antoine, VANDERMEULEN, Rob, LEI, Yunwen, ENDERS, Matthias, BORTH, Damian, KLOFT, Marius
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2021
• Subjects: Convolutional Neural Networks; Interpretability; Computer Vision.; Artificial Intelligence and Robotics; Graphics and Human Computer Interfaces
• Online Access: https://ink.library.smu.edu.sg/sis_research/7206; https://ink.library.smu.edu.sg/context/sis_research/article/8209/viewcontent/Interpretable_CNNs.pdf
• Institution: Singapore Management University

We propose a novel training methodology---Concept Group Learning (CGL)---that encourages training of interpretable CNN filters by partitioning filters in each layer into concept groups, each of which is trained to learn a single visual concept. We achieve this through a novel regularization strategy...