Variational Self-Supervised Contrastive Learning Using Beta Divergence
This addresses the problem of robust learning from noisy, unlabeled data in multi-label settings, particularly for face understanding, and is incremental as it builds on existing self-supervised and variational methods.
The paper tackled learning a discriminative semantic space from unlabeled, noisy multi-label data by proposing a variational self-supervised contrastive learning method using beta divergence, which outperformed state-of-the-art methods in accuracy across most tested scenarios.
Learning a discriminative semantic space using unlabelled and noisy data remains unaddressed in a multi-label setting. We present a contrastive self-supervised learning method which is robust to data noise, grounded in the domain of variational methods. The method (VCL) utilizes variational contrastive learning with beta-divergence to learn robustly from unlabelled datasets, including uncurated and noisy datasets. We demonstrate the effectiveness of the proposed method through rigorous experiments including linear evaluation and fine-tuning scenarios with multi-label datasets in the face understanding domain. In almost all tested scenarios, VCL surpasses the performance of state-of-the-art self-supervised methods, achieving a noteworthy increase in accuracy.