Rethinking Positive Pairs in Contrastive Learning
This work addresses representation learning in AI by exploring a novel approach to similarity learning, potentially impacting computer vision and related domains.
The paper tackles the problem of learning visual representations from semantically distinct sample pairs, proposing SimLAP to exploit similarity in subspaces between classes, with experimental demonstration of feasibility.
The training methods in AI do involve semantically distinct pairs of samples. However, their role typically is to enhance the between class separability. The actual notion of similarity is normally learned from semantically identical pairs. This paper presents SimLAP: a simple framework for learning visual representation from arbitrary pairs. SimLAP explores the possibility of learning similarity from semantically distinct sample pairs. The approach is motivated by the observation that for any pair of classes there exists a subspace in which semantically distinct samples exhibit similarity. This phenomenon can be exploited for a novel method of learning, which optimises the similarity of an arbitrary pair of samples, while simultaneously learning the enabling subspace. The feasibility of the approach will be demonstrated experimentally and its merits discussed.