Search papers, labs, and topics across Lattice.
2
0
3
By explicitly modeling relationships between detected instances and their surrounding scene context using VLMs, InCoM-Net achieves state-of-the-art performance in HOI detection, suggesting that holistic scene understanding is crucial for accurate interaction recognition.
Forget manual labeling: STONE offers a massive, automatically-labeled dataset for off-road robot navigation, unlocking scalable training for robust 3D traversability prediction.