Junhee Lee

Curiosity-driven exploration can dramatically enhance the robustness of source term estimation in noisy environments, outperforming traditional methods.

Incorporating physical dependencies into gas source localization allows for accurate inference even in challenging indoor environments, outperforming traditional methods.

Federated learning for depth estimation can be robust even in the face of overlapping data distributions among heterogeneous robots.

VLMs can detect harassment in VR with comparable accuracy to prior methods, but with far fewer training examples and without relying on sensitive biometric data.