TY - RPRT TI - The challenge of representation learning: Improved accuracy in deep vision models does not come with better predictions of perceptual similarity AU - Fritz Günther AU - Marco Marelli AU - Marco Alessandro Petilli PY - 2023 UR - https://arxiv.org/abs/2303.07084 ID - 2303.07084 ER -