Visual Emotion Recognition on Twitter I
95.63AccuracyUniEmo
Evaluation results
| Method | Links | |
|---|---|---|
| UniEmoBackbone=ViT-L/142025.07 | 95.63 | |
| UniEmoBackbone=ViT-B/162025.07 | 94.75 | |
| UniEmoBackbone=ViT-B/322025.07 | 93.98 | |
| MVPBackbone=ViT-B/322025.07 | 92.03 | |
| CoCoOpBackbone=ViT-B/322025.07 | 91.63 | |
| PT-DPCBackbone=ViT-B/322025.07 | 90.94 | |
| SimEmotionBackbone=ViT-B/322025.07 | 89.76 | |
| CoOpBackbone=ViT-B/322025.07 | 89.05 | |
| ResNet-50Enhancement strategy=w/ NS-LF2026.04 | 82.9 | |
| ResNet-50Enhancement strategy=NS-LF2026.04 | 82.9 | |
| ResNet-50Enhancement strategy=Basic2026.04 | 81.3 | |
| ResNet-50Enhancement strategy=Basic2026.04 | 81.3 | |
| DenseNet-121Enhancement strategy=w/ NS-LF2026.04 | 79.4 | |
| DenseNet-121Enhancement strategy=NS-LF2026.04 | 79.4 | |
| DenseNet-121Enhancement strategy=Basic2026.04 | 78.9 | |
| DenseNet-121Enhancement strategy=Basic2026.04 | 78.9 | |
| VGG-16Enhancement strategy=w/ NS-LF2026.04 | 78.1 | |
| VGG-16Enhancement strategy=NS-LF2026.04 | 78.1 | |
| Avg.Enhancement strategy=w/ NS-LF2026.04 | 77 | |
| ViT-B/16Enhancement strategy=w/ NS-LF2026.04 | 76.4 | |
| ViT-B/16Enhancement strategy=NS-LF2026.04 | 76.4 | |
| Avg.Enhancement strategy=Basic2026.04 | 75.5 | |
| VGG-16Enhancement strategy=Basic2026.04 | 75.3 | |
| VGG-16Enhancement strategy=Basic2026.04 | 75.3 | |
| AlexNetEnhancement strategy=w/ NS-LF2026.04 | 75.2 | |
| AlexNetEnhancement strategy=NS-LF2026.04 | 75.2 | |
| ViT-B/16Enhancement strategy=Basic2026.04 | 74.5 | |
| ViT-B/16Enhancement strategy=Basic2026.04 | 74.5 | |
| AlexNetEnhancement strategy=Basic2026.04 | 73.8 | |
| AlexNetEnhancement strategy=Basic2026.04 | 73.8 | |
| Swin-TEnhancement strategy=w/ NS-LF2026.04 | 69.7 | |
| Swin-TEnhancement strategy=NS-LF2026.04 | 69.7 | |
| Swin-TEnhancement strategy=Basic2026.04 | 69.3 | |
| Swin-TEnhancement strategy=Basic2026.04 | 69.3 |