Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Facial Expression Recognition research benchmarks – Page 2 · SOTA2 Research

Facial Expression Recognition

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

MAFW (5-fold cross-validation)

ViT-H FlashLite

48.22UAR

17

Jun 30, 2026

76.53Accuracy

17

Feb 26, 2026

89.6Accuracy

17

Feb 26, 2026

92.21Accuracy

16

Apr 1, 2026

AffectNet 7 cls (val)

67.49Accuracy

15

Feb 26, 2026

62.16Accuracy

15

Feb 26, 2026

SFEW 2.0 (test)

61.6Accuracy

15

Feb 26, 2026

82.32Accuracy (Unknown Class Only)

14

May 19, 2026

99.69Accuracy

14

Feb 26, 2026

69.8WAR

13

Mar 31, 2026

80.8WAR

13

Mar 31, 2026

81.5WAR

13

Mar 31, 2026

EMER 7-class (test)

51.18WAR

13

Feb 26, 2026

EMER 3-class (test)

68.1WAR

13

Feb 26, 2026

RAF-DB Compound

AMDiT (cnn = cp)

68.18Accuracy

12

Apr 1, 2026

69.2WAR

12

Mar 24, 2026

StressID (target)

81.5WAR

12

Mar 24, 2026

BioVid Part A (target)

81Weighted Accuracy Rate (WAR)

12

Mar 24, 2026

RAF-DB Pose > 45

94.44Accuracy

12

Feb 26, 2026

RAF-DB Pose > 30

95.35Accuracy

12

Feb 26, 2026

FERPlus basic emotion (test)

90.62Top-1 Accuracy

12

Feb 26, 2026

FER Private 2013 (test)

73.475Accuracy

12

Feb 26, 2026

48.2Accuracy

11

Jul 7, 2026

Ava256-Scan (test)

45UAR

11

Jul 7, 2026

AffectNet 10000 labels

60.12Accuracy

11

Feb 26, 2026

Page 2 of 5

Popular datasets

AffectNet CK+FERplus RAF-DB FER

Follow for updates

Join SOTA2 Contact