unlabeled data (2) 썸네일형 리스트형 [논문리뷰] FixMatch: Simplifiying Semi-Supervised Learning with Consistency and Confidence 논문 링크 : https://arxiv.org/abs/2001.07685 FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence Semi-supervised learning (SSL) provides an effective means of leveraging unlabeled data to improve a model's performance. In this paper, we demonstrate the power of a simple combination of two common SSL methods: consistency regularization and pseudo-label arxiv.org 해당 논문은 Semi.. [논문리뷰] Regularization With Stochastic Transformation and Perturbations for Deep Semi-Supervised Learning 논문 링크 : https://proceedings.neurips.cc/paper/2016/hash/30ef30b64204a3088a26bc2e6ecf7602-Abstract.html Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning Requests for name changes in the electronic proceedings will be accepted with no questions asked. However name changes may cause bibliographic tracking issues. Authors are asked to consider this ca.. 이전 1 다음
