Research
Research
Embedded Files
Our lab is dedicated to building trustworthy AI systems with safety and reliability at the forefront.
While AI is increasingly integrated into various industries such as healthcare and autonomous driving, it remains limited due to concerns about reliability and safety. Our lab aims to develop AI systems that not only perform well but also have qualities of robustness, interpretability, integrity, and fairness. Then, based on the foundational research, we applied the techniques to various applications such as medical AI, autonomous driving. The following are our main research topics, and you can find our publications here.
Design models and systems that maintain performance across diverse scenarios and data distribution shifts, ensuring dependable operation even under unforeseen conditions.
[Key Words]
Domain Generalization / Adaptation, Adversarial Robustness, Robustness Evaluation
Develop methods to transparently explain model decisions, enabling trust for both users and developers and facilitating reliability.
[Key Words]
Multimodal Reasoning, Neuron Interpretation
Ensure data handling, model updates, and deployment pipelines maintain integrity and enhance security, guarding against tampering and drift.
[Key Words]
Data Privacy, Data Watermarking, Trustworthy Data Construction
Page updated
Google Sites
Report abuse