Didier Stricker

Reconstructing realistic hand-object interactions from video just got an order of magnitude faster, thanks to a novel Gaussian Splatting approach that ensures physical consistency.

Skip the costly training and go straight to open-vocabulary 3D reasoning with ReLaGS, which builds a 3D semantic scene graph from language-distilled Gaussians.

Shrinking a leading 3D hand mesh reconstruction model by 65% yields a 1.5x speedup with minimal accuracy loss, unlocking real-time performance on resource-constrained devices.

Unlock scalable CAD generation from unannotated 3D meshes with DreamCAD, a framework that directly produces editable BREPs from point-level supervision, outperforming existing methods and achieving over 75% user preference.

Finally, a large-scale dataset bridges the gap between musculoskeletal simulation and computer vision, enabling data-driven learning of vertebral kinematics from subtle posture variations.