1 paper
Erfan Mirhaji, Nikhil Krishnaswamy, Jill Zarestky +3
Optimizing vision models for accuracy does not force the model to learn and utilize human-understandable and domain-specific concepts and rules. We propose explaining image classif…