Robust and Rotation-Equivariant Contrastive Learning.
Sign Up to like & getrecommendations! Published in 2023 at "IEEE transactions on neural networks and learning systems"
DOI: 10.1109/tnnls.2023.3243258
Abstract: Contrastive learning (CL) methods achieve great success by learning the invariant representation from various transformations. However, rotation transformations are considered harmful to CL and are rarely used, which results in failure when the objects show… read more here.
Keywords: rotation equivariant; rotation; refosnet; unseen orientations ... See more keywords