Understanding and Mitigating Human-Labelling Errors in Supervised Contrastive Learning

arXiv:2403.06289v1 Announce Type: cross
Abstract: Human-annotated vision datasets inevitably contain a fraction of human mislabelled examples. While the detrimental effects of such mislabelling on supervised learning are well-researched, their influence on Supervised Contrastive Learning (SCL) remains largely unexplored. In this paper, we show that human-labelling errors not only differ significantly from synthetic label errors, but also pose unique challenges in SCL, different to those in traditional supervised learning methods. Specifically, our results indicate they adversely impact the learning process in …

abstract arxiv cs.ai cs.cv cs.lg datasets effects errors examples human influence labelling paper show supervised learning type understanding vision