Representation Learning on Visual-Symbolic Graphs for Video Understanding

Benchmark Model Rank Results
action-detection-on-charadesI3D + biGRU + VS-ST-MPNNmAP: 23.7