Yuenan Hou

PVCap achieves a remarkable 11.41% improvement over the state-of-the-art in 3D dense captioning by leveraging innovative data augmentation and a voxel-based architecture.

Knowledge distillation can now harness the power of multi-modal models and model snapshots, achieving top-tier segmentation performance without the typical computational overhead.

SEGA3D achieves an impressive 8.3 mIoU improvement over previous methods, redefining the standards for 3D vision-language segmentation.

MATE achieves a 4.75% boost in robotic manipulation success rates under data scarcity by leveraging specialized expert routing across heterogeneous inputs.