Search papers, labs, and topics across Lattice.
1
0
2
Fine-tuning vision-language models in pixel space leads to a dramatic boost in navigation performance, outperforming traditional methods reliant on 3D information.
