Byungjun Kim

HOPE achieves accurate pressure estimation from monocular videos, enabling robust predictions of hand-object interactions without the need for specialized sensors or extensive labeled data.

Finally, a dataset that directly compares human and robot hand dexterity on the same objects, capturing both successes and failures.

Identity drift and garment distortion in virtual try-on videos? Vanast's unified framework and synthetic triplet supervision produce high-fidelity, identity-consistent animations across diverse garments.