Facial Expression Recognition on AffectNet 7-way (test)
72.93AccuracyResEmoteNet
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| ResEmoteNet2024.09 | 72.93 | — | — | — | — | — | — | — | |
| C MT EmoAffectNet2024.09 | 69.4 | — | — | — | — | — | — | — | |
| QCS2024.11 | 67.94 | — | — | — | — | — | — | — | |
| DCS2024.11 | 67.66 | — | — | — | — | — | — | — | |
| S2D2024.09 | 67.62 | — | — | — | — | — | — | — | |
| S2Dyear=2024, Pre-trained=AffectNet-82024.11 | 67.62 | — | — | — | — | — | — | — | |
| BTNYear=20242024.07 | 67.6 | — | — | — | — | — | — | — | |
| POSTER++Year=20232024.07 | 67.49 | — | — | — | — | — | — | — | |
| POSTER++year=20232024.11 | 67.49 | — | — | — | — | — | — | — | |
| POSTERYear=ICCV 20232024.07 | 67.31 | — | — | — | — | — | — | — | |
| POSTERyear=20232024.11 | 67.31 | — | — | — | — | — | — | — | |
| PF-VITbackbone=ViT-B, pre-training data=AffectNet-270K, venue=TCSS'242022.07 | 67.23 | — | — | — | — | — | — | — | |
| APVit2024.09 | 66.91 | — | — | — | — | — | — | — | |
| APVITyear=20222024.11 | 66.91 | — | — | — | — | — | — | — | |
| Emotion-GCNBackbone=DenseNet2021.06 | 66.46 | — | — | — | — | — | — | — | |
| TANyear=20232024.11 | 66.45 | — | — | — | — | — | — | — | |
| MRANbackbone=Hybrid ViT, pre-training data=MS-Celeb-1M, venue=TCSVT'232022.07 | 66.31 | — | — | — | — | — | — | — | |
| TransFERYear=ICCV 20212024.07 | 66.23 | — | — | — | — | — | — | — | |
| LDLVAbackbone=ResNet-50, pre-training data=ImageNet1K, venue=WACV'232022.07 | 66.23 | — | — | — | — | — | — | — | |
| TransFERbackbone=Hybrid ViT, pre-training data=MS-Celeb-1M, venue=ICCV'212022.07 | 66.23 | — | — | — | — | — | — | — | |
| TransFERyear=20212024.11 | 66.23 | — | — | — | — | — | — | — | |
| FRA (FT)Text supervision=false, Evaluation protocol=Fine-tuning2024.03 | 66.16 | — | — | — | — | — | — | — | |
| FRAArch.=R50 [21], Params.=24M, Pre-training Dataset=VGGFace2 [6], Pre-training Scale=3.3M, Pre-training Supervision=face image2024.03 | 66.16 | — | — | — | — | — | — | — | |
| MAE-VITbackbone=ViT-B, pre-training data=AffectNet-270K, venue=TCSS'242022.07 | 66.09 | — | — | — | — | — | — | — | |
| ExpLLMYear=ours2024.09 | 65.93 | — | — | — | — | — | — | — | |
| FRA (EAC)Text supervision=false, Evaluation protocol=EAC2024.03 | 65.85 | — | — | — | — | — | — | — | |
| FRAYear=CVPR'242024.09 | 65.85 | — | — | — | — | — | — | — | |
| OursBackbone=ResNet-18, Pre-trained=true, Conference=NeurIPS'232023.10 | 65.73 | 86.2 | 59 | 64.2 | 66.5 | 57.8 | 64.5 | 61.9 | |
| DANYear=20212024.07 | 65.69 | — | — | — | — | — | — | — | |
| CLEFText supervision=true2024.03 | 65.66 | — | — | — | — | — | — | — | |
| CLEFYear=ICCV'232024.09 | 65.66 | — | — | — | — | — | — | — | |
| BYOLText supervision=false2024.03 | 65.65 | — | — | — | — | — | — | — | |
| MAE-VITbackbone=ViT-S, pre-training data=AffectNet-270K, venue=TCSS'242022.07 | 65.53 | — | — | — | — | — | — | — | |
| Ada-DFYear=ICASSP'232024.09 | 65.34 | — | — | — | — | — | — | — | |
| EACText supervision=false2024.03 | 65.32 | — | — | — | — | — | — | — | |
| EACYear=ECCV 20222024.07 | 65.32 | — | — | — | — | — | — | — | |
| EACYear=ECCV'222024.09 | 65.32 | — | — | — | — | — | — | — | |
| EACbackbone=ResNet-18, pre-training data=MS-Celeb-1M, venue=ECCV'222022.07 | 65.32 | — | — | — | — | — | — | — | |
| EACyear=20222024.11 | 65.32 | — | — | — | — | — | — | — | |
| DACLYear=WACV 20202024.07 | 65.2 | — | — | — | — | — | — | — | |
| ARMYear=20212024.07 | 65.2 | — | — | — | — | — | — | — | |
| EACBackbone=ResNet-18, Pre-trained=true, Conference=ECCV'222023.10 | 65.17 | 91.4 | 64.5 | 65.7 | 66.3 | 61.6 | 60.9 | 45.8 | |
| FG-AGRbackbone=Hybrid ViT, pre-training data=MS-Celeb-1M, venue=TCSVT'232022.07 | 64.91 | — | — | — | — | — | — | — | |
| FaRLText supervision=true2024.03 | 64.85 | — | — | — | — | — | — | — | |
| FaRLArch.=ViT-B/16 [17], Params.=86M, Pre-training Dataset=LAION-FACE [72], Pre-training Scale=20M, Pre-training Supervision=face image + text2024.03 | 64.85 | — | — | — | — | — | — | — | |
| VTFFYear=TAC'212024.09 | 64.8 | — | — | — | — | — | — | — | |
| MVTbackbone=DeiT-S [44], pre-training data=ImageNet1K, venue=Arxiv'212022.07 | 64.57 | — | — | — | — | — | — | — | |
| AMP-Netbackbone=ResNet-34, pre-training data=VGGFace2, venue=TCSVT'222022.07 | 64.54 | — | — | — | — | — | — | — | |
| MA-NetYear=TIP 20212024.07 | 64.53 | — | — | — | — | — | — | — | |
| MAE-VITbackbone=ViT-T, pre-training data=AffectNet-270K, venue=TCSS'242022.07 | 64.25 | — | — | — | — | — | — | — | |
| Meta-Face2ExpYear=CVPR 20222024.07 | 64.23 | — | — | — | — | — | — | — | |
| Face2Expyear=20222024.11 | 64.23 | — | — | — | — | — | — | — | |
| Siamese2021.06 | 64 | — | — | — | — | — | — | — | |
| KTNText supervision=false2024.03 | 63.97 | — | — | — | — | — | — | — | |
| KTNYear=TIP 20202024.07 | 63.97 | — | — | — | — | — | — | — | |
| PSRYear=CVPR 20202024.07 | 63.77 | — | — | — | — | — | — | — | |
| EfficientFaceYear=AAAI 20212024.07 | 63.7 | — | — | — | — | — | — | — | |
| SCNYear=CVPR'202024.09 | 63.4 | — | — | — | — | — | — | — | |
| CNNs and BOVW + local SVMData Augmentation (aug.)=true2018.04 | 63.31 | — | — | — | — | — | — | — | |
| CNNs and BOVW + global SVM2021.06 | 63.31 | — | — | — | — | — | — | — | |
| CNNs and BOVW + global SVMData Augmentation (aug.)=false2018.04 | 63.2 | — | — | — | — | — | — | — | |
| DMUEYear=CVPR 20212024.07 | 63.11 | — | — | — | — | — | — | — | |
| CNNs and BOVW + local SVMData Augmentation (aug.)=false2018.04 | 62.94 | — | — | — | — | — | — | — | |
| CNNs and BOVW + global SVMData Augmentation (aug.)=true2018.04 | 62.91 | — | — | — | — | — | — | — | |
| Arm2024.09 | 62.5 | — | — | — | — | — | — | — | |
| IPD-FERbackbone=ResNet-18, pre-training data=CASIA-WebFace [53], venue=TAFFC'222022.07 | 62.23 | — | — | — | — | — | — | — | |
| Hua et al.Data Augmentation (aug.)=true2018.04 | 62.11 | — | — | — | — | — | — | — | |
| Emotion-GCNBackbone=BReG-NeXt2021.06 | 61.94 | — | — | — | — | — | — | — | |
| OADN2021.06 | 61.89 | — | — | — | — | — | — | — | |
| VTFFYear=TAC 20212024.07 | 61.85 | — | — | — | — | — | — | — | |
| CAKE2021.06 | 61.7 | — | — | — | — | — | — | — | |
| RULBackbone=ResNet-18, Pre-trained=true, Conference=NeurIPS'212023.10 | 61.56 | 90.5 | 62.4 | 64.7 | 69.3 | 60.8 | 49 | 34.2 | |
| Facial Motion Prior Network2021.06 | 61.52 | — | — | — | — | — | — | — | |
| RULText supervision=false2024.03 | 61.43 | — | — | — | — | — | — | — | |
| RULYear=NIPS'212024.09 | 61.43 | — | — | — | — | — | — | — | |
| LEWELText supervision=false2024.03 | 61.2 | — | — | — | — | — | — | — | |
| Multi-task+CCCBackbone=BReG-NeXt2021.06 | 61.14 | — | — | — | — | — | — | — | |
| MCFText supervision=false2024.03 | 60.98 | — | — | — | — | — | — | — | |
| PCLText supervision=false2024.03 | 60.77 | — | — | — | — | — | — | — | |
| PCLYear=CVPR'232024.09 | 60.77 | — | — | — | — | — | — | — | |
| BBNBackbone=ResNet-18, Pre-trained=true, Conference=CVPR'202023.10 | 60.76 | 87 | 57.1 | 66.8 | 58.3 | 54.9 | 71.1 | 30.1 | |
| Single-taskBackbone=BReG-NeXt2021.06 | 60.49 | — | — | — | — | — | — | — | |
| SCNbackbone=ResNet-18, pre-training data=Ms-Celeb-1M, venue=CVPR'202022.07 | 60.23 | — | — | — | — | — | — | — | |
| Bulat et al.Text supervision=false2024.03 | 60.2 | — | — | — | — | — | — | — | |
| LDL-ALSGYear=CVPR 20202024.07 | 59.35 | — | — | — | — | — | — | — | |
| gACNN2021.06 | 58.78 | — | — | — | — | — | — | — | |
| FRA (LP)Text supervision=false, Evaluation protocol=Linear probe2024.03 | 57.38 | — | — | — | — | — | — | — | |
| IPA2LT2021.06 | 57.31 | — | — | — | — | — | — | — | |
| SEResNet2024.09 | 56.54 | — | — | — | — | — | — | — | |
| Li et al.Data Augmentation (aug.)=false2018.04 | 55.33 | — | — | — | — | — | — | — | |
| SCNBackbone=ResNet-18, Pre-trained=true, Conference=CVPR'202023.10 | 53.3 | 95.2 | 82.7 | 44.2 | 56.3 | 35.8 | 38 | 20.9 |