Enhancing Semi-Supervised Learning via Representative and Diverse Sample Selection

Semi-Supervised Learning (SSL) has become a preferred paradigm in many deep learning tasks, which reduces the need for human labor. Previous studies primarily focus on effectively utilising the labelled and unlabeled data to improve performance. However, we observe that how to select samples for labelling also significantly impacts performance, particularly under extremely low-budget settings. The sample selection task in SSL has been under-explored for a long time. To fill in this gap, we propose a Representative and Diverse Sample Selection approach (RDSS). By adopting a modified Frank-Wolfe algorithm to minimise a novel criterion $\alpha$-Maximum Mean Discrepancy ($\alpha$-MMD), RDSS samples a representative and diverse subset for annotation from the unlabeled data. We demonstrate that minimizing $\alpha$-MMD enhances the generalization ability of low-budget learning. Experimental results show that RDSS consistently improves the performance of several popular SSL frameworks and outperforms the state-of-the-art sample selection approaches used in Active Learning (AL) and Semi-Supervised Active Learning (SSAL), even with constrained annotation budgets. Our code is available at [RDSS](https: //github. com/YanhuiAILab/RDSS).

Authors

• Qian Shao Singapore Management University
• Jiangrui Kang
• Qiyuan Chen University of Wisconsin - Madison
• Zepeng Li Zhejiang University
• Hongxia Xu College of Computer and Cyber Security, Hebei Normal University, Hebei 050024, China Hebei Provincial Engineering Research Center for Supply Chain Big Data Analytics & Data Security, Hebei 050024, China Hebei Provincial Key Laboratory of Network and Information Security, Hebei 050024, China
• Yiwen Cao South China University of Technology
• Jiajuan Liang
• Jian Wu State Key Laboratory of Transvascular Implantation Devices of The Second Affiliated Hospital, Zhejiang University Zhejiang Key Laboratory of Medical Imaging Artificial Intelligence

Keywords

No keywords are indexed for this paper.