Abstract

Assuming unknown classes could be present during classification, the open set recognition (OSR) task aims to classify an instance into a known class or reject it as unknown. In this paper, we use a two-stage training strategy for OSR problems. In the first stage, we introduce a self-supervised feature decoupling method that finds the content features of the input samples from the known classes. Specifically, our feature decoupling approach learns a representation that can be split into content features and transformation features. In the second stage, we fine-tune the content features with the class labels. The fine-tuned content features are then used for the OSR problems. To measure representation quality, we introduce intra-inter ratio (IIR). Our experimental results indicate that our proposed self-supervised approach outperforms others in image and malware OSR problems. Also, our analyses indicate that IIR is correlated with and can explain OSR performance.

About the Speaker

Philip Chan is an Associate Professor of Computer Science at Florida Institute of Technology. His main research interests include machine learning, data mining, and distributed computing. He received his PhD, MS, and BS in Computer Science from Columbia University, Vanderbilt University, and Texas State University respectively. He is an Associate Editor for ACM Transactions on Knowledge Discovery from Data (TKDD), Transactions on Machine Learning Research (TMLR), and the Knowledge and Information Systems (KAIS) journal.