List your company
ResearchFormerly WizWand

Action Recognition on UTD-MHAD (val)

94.5AccuracyPoseMap
24.518442.686760.85579.0233Sep 27, 2021
Updated 8mo ago

Evaluation results

MethodLinks
PoseMapApproach=PoseMap2021.0994.5
Fusion-GCNApproach=Skeleton + IMU (Center)2021.0994.42
Fusion-GCNApproach=Skeleton + IMU (Wrist/Hip)2021.0994.07
Gimme SignalsApproach=Gimme Signals2021.0993.33
Fusion-GCNApproach=Skeleton + IMU (Center + Add. Edges)2021.0993.26
Fusion-GCNApproach=Skeleton + IMU (Wrist/Hip + Add. Edges)2021.0993.26
MCRLApproach=MCRL2021.0993.02
Fusion-GCNApproach=Skeleton + IMU + RGB Encoder R(2+1)D2021.0992.85
Fusion-GCNApproach=Skeleton + IMU + RGB Encoder R-182021.0992.33
SkeletonApproach=Skeleton2021.0992.32
Fusion-GCNApproach=Skeleton + RGB Encoder R(2+1)D2021.0991.62
Fusion-GCNApproach=Skeleton + IMU (Channel Fusion)2021.0990.29
Fusion-GCNApproach=Skeleton + RGB Encoder R-182021.0989.83
Fusion-GCNApproach=Skeleton + IMU + RGB Patch Features R-182021.0978.9
Fusion-GCNApproach=Skeleton + RGB Patch Features R-182021.0973.49
R(2+1)DApproach=R(2+1)D2021.0961.63
Fusion-GCNApproach=Skeleton + RGB Patch Features R-18 (no MLP)2021.0944.6
RGB Patch Features R-18Approach=RGB Patch Features R-182021.0927.67
RGB Encoder R-18Approach=RGB Encoder R-182021.0927.21