List your company
ResearchFormerly WizWand

Violence Detection on Crowd Violence

96.33AccuracyDIFEM
80.698884.756988.81592.8731Nov 14, 2019Sep 17, 2020Jul 22, 2021May 26, 2022Mar 30, 2023Feb 1, 2024Dec 6, 2024
Updated 7mo ago

Evaluation results

MethodLinks
DIFEMclassifier=Random Forest, protocol=5-fold cross validation2024.1296.33
Ullah et al.2024.1296
Hachiuma et al.2024.1294.7
ConvLSTMType=Deep-Learning Based2019.1194.57
SPIL2024.1294.5
LHOG+LOFType=Hand-Crafted Features2019.1194.31
LHOG+LOF2024.1294.31
3D ConvNetType=Deep-Learning Based2019.1194.3
Garcia-Cobo et al.2024.1294.3
DIFEMclassifier=AdaBoost, protocol=5-fold cross validation2024.1294.3
DIFEMclassifier=Nearest Neighbor, protocol=5-fold cross validation2024.1294.3
DIFEMclassifier=Decision Tree, protocol=5-fold cross validation2024.1293.47
I3D(Flow only)Type=Deep-Learning Based2019.1188.89
I3D(Fusion)Type=Deep-Learning Based2019.1188.89
Flow Gated NetworkType=Deep-Learning Based2019.1188.87
Cheng et al.2024.1288.8
ViF + OViF2024.1288
DGCNNevaluation_source=Reported in [24]2024.1287.4
I3D(RGB only)Type=Deep-Learning Based2019.1186.67
ViF2024.1285
C3DType=Deep-Learning Based2019.1184.44
I3Devaluation_source=Reported in [24]2024.1283.4
MoWLD+BOWType=Hand-Crafted Features2019.1182.56
MoWLD+BoW2024.1282.56
TSNevaluation_source=Reported in [24]2024.1281.5
ViFType=Hand-Crafted Features2019.1181.3