List your company
ResearchFormerly WizWand

Facial Expression Recognition on AFEW 8.0 (test)

55.17AccuracyNoisy Student Training
44.14647.00849.8752.732Jun 29, 2019Sep 4, 2019Nov 10, 2019Jan 17, 2020Mar 24, 2020May 30, 2020Aug 6, 2020
Updated 8mo ago

Evaluation results

MethodLinks
Noisy Student Trainingiter. training=true2020.0855.17
VGG-Face + BLSTMYear=20182020.0853.91
Noisy Student Trainingiter. training=false2020.0852.49
DenseNet-161Model type=Static2019.0651.44
DenseNet-161Year=20182020.0851.44
FANModel type=Static2019.0651.18
FANYear=20192020.0851.18
FAN w/o Relation-attentionModel type=Static2019.0650.92
ResNet-18Year=20182020.0849.7
VGG-FaceYear=20192020.0849
Score fusion (baseline)Model type=Static2019.0648.82
VGGFace + Undirectional LSTMModel type=Dynamic2019.0648.6
VGG-Face + LSTMYear=20172020.0848.6
DSN-VGGFaceModel type=Static2019.0648.04
DSN-VGGFaceYear=20182020.0848.04
DSN-HoloNetModel type=Static2019.0646.47
DSN-HoloNetYear=20172020.0846.47
CNN-RNNModel type=Dynamic2019.0645.43
CNN-RNNYear=20162020.0845.43
HoloNetModel type=Static2019.0644.57