Regular biography
Jinwoo Shin is a faculty member in the Department of Computer Science at KAIST. His research is associated with the Kim Jaechul Graduate School of AI. He can be contacted via email at jinwoos@kaist.ac.kr, and his profile page is available at https://sites.google.com/site/mijirim/. The provided profile page text includes his name, affiliation, and some basic information about his academic presence.
Scholar-generated biography
Jinwoo Shin is an ICT Endowed Chair Professor specializing in Machine Learning and Deep Learning. His research focuses on developing robust frameworks for detecting out-of-distribution samples and adversarial attacks, as well as improving classifier reliability through confidence calibration. He explores novelty detection using contrastive learning on distributionally shifted instances and addresses bias in classifiers by de-biasing techniques. His work also includes contrastive continual learning, representation alignment for generation, and imbalanced classification methods. Additionally, he investigates reinforcement learning, video delivery, and generalization in deep learning through techniques like network randomization and pseudo-label refinement. His contributions span both theoretical and applied aspects of machine learning, emphasizing robustness and adaptability.