VSGNet

Reported on 4 benchmarks across 1 task · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision4 results

Human-Object Interaction DetectiononV-COCO
AP(S1)· 2020-03-11
51.76
best: 72.1 (RLIPv2)
SOTA
VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions arXiv:2003.05541
Human-Object Interaction DetectiononV-COCO
AP(S2)· 2020-03-11
57
best: 74.1 (RLIPv2)
SOTA
VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions arXiv:2003.05541
Human-Object Interaction DetectiononV-COCO
Time Per Frame(ms)· 2020-03-11
312
best: 513 (TIN (Interactiveness))
VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions arXiv:2003.05541
Human-Object Interaction DetectiononHICO-DET
mAP· 2020-03-11
19.8
best: 46.49 (Ours (PViC+))
VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions arXiv:2003.05541