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.
@article{arxiv.2410.18200,
title = {Rethinking Positive Pairs in Contrastive Learning},
author = {Jiantao Wu and Sara Atito and Zhenhua Feng and Shentong Mo and Josef Kitler and Muhammad Awais},
journal= {arXiv preprint arXiv:2410.18200},
year = {2025}
}