Facial Action Unit Detection on DISFA
65.4F1 (AU 1)MDHR
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| MDHRModeling Approach=Spatio-temporal, Backbone=Swin-B2024.04 | 65.4 | 60.2 | 75.2 | 50.2 | 52.4 | 74.3 | 93.7 | 58.2 | 66.2 | |
| MDHR2026.02 | 65.4 | 60.2 | 75.2 | 50.2 | 52.4 | 74.3 | 93.7 | 58.2 | 66.2 | |
| CLEF2026.02 | 64.3 | 61.8 | 68.4 | 49 | 55.2 | 72.9 | 89.9 | 57 | 64.8 | |
| AAR2026.02 | 62.4 | 53.6 | 71.5 | 39 | 48.8 | 76.1 | 91.3 | 70.6 | 64.2 | |
| MDHRModeling Approach=Spatio-temporal, Backbone=ResNet-502024.04 | 61.4 | 57.7 | 70.9 | 57.1 | 48.3 | 75.7 | 91.5 | 56.7 | 64.9 | |
| MGRR-Net2026.02 | 61.3 | 62.9 | 75.8 | 48.7 | 53.8 | 75.5 | 94.3 | 73.1 | 68.2 | |
| VL-FAU2026.02 | 60.9 | 56.4 | 74 | 46.3 | 60.8 | 72.4 | 94.3 | 66.5 | 66.5 | |
| HiVABackbone=Swin Transformer Base2026.02 | 60.6 | 58.4 | 75.4 | 51 | 61.2 | 74.8 | 93.9 | 63.8 | 67.4 | |
| AUNet2026.02 | 60.3 | 59.1 | 69.8 | 48.4 | 53 | 79.7 | 93.5 | 64.7 | 66.1 | |
| WSRTLModeling Approach=Spatio-temporal2024.04 | 57.3 | 51.8 | 74.3 | 49.8 | 44.8 | 79.3 | 94.6 | 64.6 | 64.6 | |
| FAN-TransModeling Approach=Static AU relationship modeling2024.04 | 56.4 | 50.2 | 68.6 | 49.2 | 57.6 | 75.6 | 93.6 | 58.8 | 63.8 | |
| SEV-Net2026.02 | 55.3 | 53.1 | 61.5 | 53.6 | 38.2 | 71.6 | 95.7 | 41.5 | 58.8 | |
| ME-GraphAUModeling Approach=Static AU relationship modeling2024.04 | 54.6 | 47.1 | 72.9 | 54 | 55.7 | 76.7 | 91.1 | 53 | 63.1 | |
| ME-GraphAU2026.02 | 54.6 | 47.1 | 72.9 | 54 | 55.7 | 76.7 | 91.1 | 53 | 63.1 | |
| HSTR-NetModeling Approach=Spatio-temporal2024.04 | 54.3 | 50.8 | 70.1 | 66.6 | 59.6 | 68 | 97.9 | 69.8 | 62.9 | |
| PCLTraining paradigm=Self-supervised2022.11 | 53.8 | 44.9 | 58.1 | 37.2 | 53.2 | 73.1 | 86.5 | 31.3 | 54.8 | |
| KSModeling Approach=Spatio-temporal2024.04 | 53.8 | 59.9 | 69.2 | 54.2 | 50.8 | 75.8 | 92.2 | 46.8 | 62.8 | |
| SMA-NetModeling Approach=Static image-based2024.04 | 53.4 | 54.2 | 64 | 57 | 47 | 76.6 | 92 | 55.2 | 64.2 | |
| STRALModeling Approach=Spatio-temporal2024.04 | 52.2 | 47.4 | 68.9 | 47.8 | 56.7 | 72.5 | 91.3 | 67.6 | 63 | |
| PIAP2026.02 | 50.2 | 51.8 | 71.9 | 50.6 | 54.5 | 79.7 | 94.1 | 57.2 | 63.8 | |
| CISNet2026.02 | 48.8 | 50.4 | 78.9 | 51.9 | 47.1 | 80.1 | 95.4 | 65 | 64.7 | |
| FAUDTModeling Approach=Static AU relationship modeling2024.04 | 46.1 | 48.6 | 72.8 | 56.7 | 50 | 72.1 | 90.8 | 55.4 | 61.5 | |
| Jacob et al.2026.02 | 46.1 | 48.6 | 72.8 | 56.7 | 50 | 72.1 | 90.8 | 55.4 | 61.5 | |
| SRERLModeling Approach=Static AU relationship modeling2024.04 | 45.7 | 47.8 | 59.6 | 47.1 | 45.6 | 73.5 | 84.3 | 43.6 | 55.9 | |
| SRERL2026.02 | 45.7 | 47.8 | 59.6 | 47.1 | 45.6 | 73.5 | 84.3 | 43.6 | 55.9 | |
| ATF2026.02 | 45.2 | 39.7 | 47.1 | 48.6 | 32 | 55 | 86.4 | 39.2 | 49.2 | |
| ARLModeling Approach=Static image-based2024.04 | 43.9 | 42.1 | 63.6 | 41.8 | 40 | 76.2 | 95.2 | 66.8 | 58.7 | |
| ARL2026.02 | 43.9 | 42.1 | 63.6 | 41.8 | 40 | 76.2 | 95.2 | 66.8 | 58.7 | |
| JAA-NetTraining paradigm=Supervised2022.11 | 43.7 | 46.2 | 56 | 41.4 | 44.7 | 69.6 | 88.3 | 58.4 | 56 | |
| JAA-NetModeling Approach=Static image-based2024.04 | 43.7 | 46.2 | 56 | 41.4 | 44.7 | 69.6 | 88.3 | 58.4 | 56 | |
| JAA-Net2026.02 | 43.7 | 46.2 | 56 | 41.4 | 44.7 | 69.6 | 88.3 | 58.4 | 56 | |
| UGN2026.02 | 43.3 | 48.1 | 63.4 | 49.5 | 48.2 | 72.9 | 90.8 | 59 | 60 | |
| DSIN2026.02 | 42.4 | 39 | 68.4 | 28.6 | 46.8 | 70.8 | 90.4 | 42.2 | 53.6 | |
| EAC-NetTraining paradigm=Supervised2022.11 | 41.5 | 26.4 | 66.4 | 50.7 | 80.5 | 89.3 | 88.9 | 15.6 | 48.5 | |
| EAC-NetModeling Approach=Static image-based2024.04 | 41.5 | 26.4 | 66.4 | 50.7 | 80.5 | 89.3 | 88.9 | 15.6 | 48.5 | |
| EAC-Net2026.02 | 41.5 | 26.4 | 66.4 | 50.7 | 80.5 | 89.3 | 88.9 | 15.6 | 48.5 | |
| SimCLRTraining paradigm=Self-supervised, Reproduced=true2022.11 | 40.5 | 46.9 | 53.8 | 33.5 | 24.9 | 74.7 | 85 | 35.6 | 49.4 | |
| SO-Net2026.02 | 33.8 | 44.5 | 70.3 | 57.6 | 39.7 | 78.2 | 86.7 | 57.3 | 58.5 | |
| FaceCycleTraining paradigm=Self-supervised, Reproduced=true2022.11 | 26.4 | 10.2 | 37.3 | 21.5 | 25 | 71.8 | 84.2 | 34.7 | 38.9 | |
| IdenNet2026.02 | 25.5 | 34.8 | 64.5 | 45.2 | 44.6 | 70.7 | 81 | 55 | 52.6 | |
| TAE2026.02 | 21.4 | 19.6 | 64.5 | 46.8 | 44 | 73.2 | 85.1 | 55.3 | 51.5 | |
| DeformAETraining paradigm=Self-supervised2022.11 | 17.6 | 12.3 | 46.7 | 43.5 | 26 | 62.7 | 64.8 | 47.6 | 40.1 | |
| DRMLTraining paradigm=Supervised2022.11 | 17.3 | 17.7 | 37.4 | 29 | 10.7 | 37.7 | 38.5 | 20.1 | 26.7 | |
| Fab-NetTraining paradigm=Self-supervised2022.11 | 15.5 | 16.2 | 43.2 | 50.4 | 23.2 | 69.6 | 72.4 | 42.4 | 41.6 | |
| TCAETraining paradigm=Self-supervised2022.11 | 15.1 | 16.2 | 50.5 | 48.7 | 23.3 | 72.1 | 72.4 | 42.4 | 45 | |
| SplitBrainTraining paradigm=Self-supervised2022.11 | 13.1 | 10.6 | 35.7 | 40.2 | 30.2 | 57.5 | 77.4 | 40.3 | 38.1 | |
| TCAETraining paradigm=Self-supervised, Reproduced=true2022.11 | 10.5 | 13.3 | 20.9 | 18.8 | 7.5 | 44.7 | 57.8 | 9.9 | 22.9 |