A new approach to training more interpretable model with additional segmentation

Sunguk Shin, Youngjoon Kim, Ji Won Yoon

Abstract: Highlights•The classification and interpretability are increased by using the guidance image.•CAM, one of interpretable machine learning, is applied to the classification model.•The performance is increased with the Gaussian Process for adjusting two losses.•The performance is increased in proportion to the number of guidance images.