List your company
ResearchFormerly WizWand

Violence Detection on Hockey Fight

99.62Accuracy3D ConvNet
70.260877.882985.50593.1271Nov 14, 2019Sep 17, 2020Jul 22, 2021May 26, 2022Mar 30, 2023Feb 1, 2024Dec 6, 2024
Updated 7mo ago

Evaluation results

MethodLinks
3D ConvNetType=Deep-Learning Based2019.1199.62
Hachiuma et al.2024.1299.5
Zhang at al.2024.1299.4
Islam et al.2024.1299
DIFEMclassifier=Random Forest, protocol=5-fold cross validation2024.1298.8
SSHAInput modality=RGB2022.0298.7
DIFEMclassifier=Nearest Neighbor, protocol=5-fold cross validation2024.1298.7
I3D(RGB only)Type=Deep-Learning Based2019.1198.5
I3DInput modality=RGB2022.0298.5
DIFEMclassifier=AdaBoost, protocol=5-fold cross validation2024.1298.5
Flow Gated NetworkType=Deep-Learning Based2019.1198
Cheng et al. ModelArchitecture=P3D2022.0298
SSHAInput modality=RGB, Localization=Disabled2022.0298
Cheng et al.2024.1298
I3D(Fusion)Type=Deep-Learning Based2019.1197.5
I3DInput modality=Two-stream2022.0297.5
ConvLSTMType=Deep-Learning Based2019.1197.1
ConvLSTM2022.0297.1
DIFEMclassifier=Decision Tree, protocol=5-fold cross validation2024.1297.1
FightNetType=Deep-Learning Based2019.1197
SSHAInput modality=Two-stream2022.0297
SPIL2024.1296.8
C3DType=Deep-Learning Based2019.1196.5
LHOG+LOFType=Hand-Crafted Features2019.1195.1
LHOG+LOF2024.1295.1
Garcia-Cobo et al.2024.1294.5
I3Devaluation_source=Reported in [24]2024.1293.4
MoWLD+BOWType=Hand-Crafted Features2019.1191.9
MoWLD+BoW2024.1291.9
HOG+HIKType=Hand-Crafted Features2019.1191.7
HOG+HIK2024.1291.7
TSNevaluation_source=Reported in [24]2024.1291.5
MOSIFT+HIKType=Hand-Crafted Features2019.1190.9
DGCNNevaluation_source=Reported in [24]2024.1290.2
HOF+HIKType=Hand-Crafted Features2019.1188.6
ViF + OViF2024.1287.5
SSHAInput modality=Optical-flow2022.0286.2
I3D(Flow only)Type=Deep-Learning Based2019.1184
I3DInput modality=Optical-flow2022.0284
ViFType=Hand-Crafted Features2019.1182.9
ViF2024.1282.9
3D CNNevaluation_source=Reported in [24]2024.1282.6
Tran et al.2024.1271.39