SSDL: Self-Supervised Domain Learning for Improved Face Recognition
This work aims to improve face recognition performance in unconstrained environments for various applications by reducing the need for expensive domain-specific annotated data, representing an incremental improvement in self-supervised learning for this domain.
This paper addresses the challenge of face recognition in unconstrained environments where variations in conditions create a domain gap between training and testing data. The authors propose a self-supervised domain learning (SSDL) scheme that trains on triplets mined from unlabelled data using an "easy-to-hard" approach, demonstrating good generalization across four benchmarks.
Face recognition in unconstrained environments is challenging due to variations in illumination, quality of sensing, motion blur and etc. An individual's face appearance can vary drastically under different conditions creating a gap between train (source) and varying test (target) data. The domain gap could cause decreased performance levels in direct knowledge transfer from source to target. Despite fine-tuning with domain specific data could be an effective solution, collecting and annotating data for all domains is extremely expensive. To this end, we propose a self-supervised domain learning (SSDL) scheme that trains on triplets mined from unlabelled data. A key factor in effective discriminative learning, is selecting informative triplets. Building on most confident predictions, we follow an "easy-to-hard" scheme of alternate triplet mining and self-learning. Comprehensive experiments on four different benchmarks show that SSDL generalizes well on different domains.