Search papers, labs, and topics across Lattice.
4
0
5
3
Navigation models trained in Image2Sim's synthetic environments outperform traditional methods, achieving zero-shot transfer to real-world settings.
NoPA retains rich 3D geometric information while achieving real-time scene graph generation, outperforming traditional methods that sacrifice detail for speed.
DART boosts reasoning accuracy by up to 22.5 points while slashing thinking token usage by over 50%, all without requiring labeled training data.
Forget offline optimization – EmbodiedSplat brings real-time, open-vocabulary semantic 3D scene understanding to embodied agents by reconstructing and understanding scenes from streaming images.
