Visual Sentiment Classification on DeepSent five-agree ⟨σ5, P2⟩
95.6AccuracyMLLMsent
Evaluation Results
| Method | Links | |
|---|---|---|
| MLLMsentTraining dataset=PerceptSent, Fine-tuned on DeepSent=false, Evaluation mode=Cross-dataset generalization2025.08 | 95.6 | |
| DenseNet169Architecture=Deep CNN2025.08 | 88.3 | |
| InceptionV3Architecture=Deep CNN2025.08 | 87.7 | |
| ResNet50Architecture=Deep CNN2025.08 | 86.5 | |
| Campos et al.2025.08 | 83 | |
| VGG16Architecture=Deep CNN2025.08 | 79.4 | |
| You et al.2025.08 | 74.7 | |
| LCHType=Hand-crafted2025.08 | 71 | |
| GCHType=Hand-crafted2025.08 | 68.4 |