Search papers, labs, and topics across Lattice.
Affiliation:
5
0
7
Agents can now learn to design and critique 3D scenes like human designers, achieving state-of-the-art quality through complementary learning.
TAMP-Nav aligns embodied navigation with VLMs' 2D capabilities, achieving a 66.2% success rate while drastically improving efficiency.
Image-generation models can outperform text-output VLMs in spatial tasks when evaluated through visual answers, revealing a critical shift in how we assess spatial intelligence in AI.
Landmark bias can lead to significant inaccuracies in geo-localization, but HoloGeo effectively mitigates this issue through evidence-driven reasoning, outperforming existing models.