J Heylen, M De Wolf, B Dawagne… - Proceedings of the …, 2021 - openaccess.thecvf.com
… This results in the network learning an internal representation of the camera intrinsics, by
linking position and scale in the image to real world depth. We show that, for such …