Search papers, labs, and topics across Lattice.
1
0
3
Decomposing accident understanding into temporal, semantic, and spatial components leads to a remarkable boost in zero-shot reasoning accuracy for surveillance video analysis.