List your company

Grounded Situation Recognition on SWiG 1.0 (test)

58.19Top-1 Verb AccClipSitu XTF
31.30638.285545.26552.2445Jul 2, 2023

Evaluation results

MethodLinks
ClipSitu XTFBackbone=ViT-L14, Input Resolution=336px2023.0758.1947.2329.7340.0115.0385.6968.4241.4249.7825.2278.5245.3154.3633.2
CLIP-Event2023.0745.633.120.121.610.6—————————
CoFormer2023.0744.6635.9822.2229.0512.2173.3157.7633.9846.2518.3775.9541.8760.1122.12
SituFormer2023.0744.235.2421.8629.2213.4171.2155.7533.274620.175.8542.1361.8924.89
Kernel GraphNet2023.0743.2735.4119.38——68.7255.6230.29——72.9242.35——
GSRTR2023.0740.6332.1519.2825.4910.169.8154.1331.0142.515.8874.113957.4519.67
JSL2023.0739.9431.4418.8724.869.6667.651.8829.3940.614.7273.2137.8256.5718.45
ISL2023.0739.3630.0918.6222.737.7265.5150.1628.4736.611.5672.4237.152.1914.58
CAQBackbone=RE-VGG2023.0738.1930.2318.47——65.0550.2128.93——73.4138.52——
GraphNet2023.0736.7227.5219.25——61.945.3929.96——69.1641.36——
RNN w/ Fusion2023.0735.927.4516.36——63.0846.8826.06——70.2735.25——
CRF w/ DataAug2023.0734.1226.4515.51——62.5946.8825.46——70.4434.38——
CRF2023.0732.3424.6414.19——58.8842.7622.55——65.6628.96——