List your company
ResearchFormerly WizWand

Engagement Recognition on EngageNet

73.58AccuracyPriorNet
53.965659.057864.1569.2422Apr 2, 2026Apr 11, 2026Apr 20, 2026Apr 29, 2026May 8, 2026May 17, 2026May 27, 2026
Updated 4mo ago

Evaluation results

MethodLinks
PriorNetPrinciple=Prior-guided2026.0573.58
ST-GCNYear=20242026.0471.24
Ordinal ST-GCNPrinciple=Ordinal graph2026.0571.24
ST-GCNYear=2024, Ind. Features=true, Features=Facial landmarks2026.0571.24
ST-GCNPrinciple=Graph-based2026.0569.37
TransformerYear=20232026.0469.1
TransformerPrinciple=Transformer-based2026.0569.1
TransformerYear=2023, Ind. Features=true, Features=Gaze + Head Pose + AU2026.0569.1
VE-MDYear=20262026.0468.98
VE_MDYear=2025, Ind. Features=false, Features=Pred. Body+Face SR2026.0568.98
TCCT-NetYear=20242026.0468.91
TCCT-NetPrinciple=Hybrid temporal2026.0568.91
TCCT-NetYear=2024, Ind. Features=true, Features=Head Pose2026.0568.91
GLAMOR-NetYear=20242026.0468.72
GLAMOR-NetYear=2024, Ind. Features=true, Features=Facial features2026.0568.72
VLM-filteringPrinciple=VLM-assisted filtering2026.0568.16
TCNPrinciple=Temporal features2026.0567.79
LSTMPrinciple=Sequential features2026.0567.69
CNN + LSTMPrinciple=Sequential features2026.0567.51
TransformerYear=20232026.0464.45
TransformerYear=2023, Ind. Features=true, Features=Gaze + Head Pose2026.0564.45
EfficientNet + Bi-LSTMPrinciple=End-to-end2026.0558.94
EfficientNet + LSTMPrinciple=End-to-end2026.0557.57
TransformerYear=20232026.0455.45
TransformerYear=2023, Ind. Features=true, Features=Gaze2026.0555.45
ResNet + TCNPrinciple=End-to-end2026.0554.72