Image Classification on Hateful Memes
66AccuracyOpen-Flamingo MOE
Evaluation results
| Method | Links | |
|---|---|---|
| Open-Flamingo MOEText=Raw Text, ICL Tokens=256, Shots=322024.06 | 66 | |
| CLIPBackbone=ResNet-50, Evaluation Protocol=Linear probe2021.06 | 65.7 | |
| VisInContextText=Rendered Image, ICL Tokens=2048, Shots=42024.06 | 65.5 | |
| VisInContextText=Rendered Image, ICL Tokens=2048, Shots=322024.06 | 65.2 | |
| Open-Flamingo MOEText=Raw Text, ICL Tokens=256, Shots=42024.06 | 63.8 | |
| VisInContextText=Rendered Image, ICL Tokens=2048, Shots=02024.06 | 62.2 | |
| Open-Flamingo MOEText=Raw Text, ICL Tokens=256, Shots=02024.06 | 60.4 | |
| SupervisedBackbone=ResNet-50, Evaluation Protocol=Linear probe2021.06 | 56.5 | |
| EsViTBackbone=Swin-T, Evaluation Protocol=Linear probe2021.06 | 52 | |
| SupervisedBackbone=Swin-T, Evaluation Protocol=Linear probe2021.06 | 51.2 | |
| SupervisedBackbone=ResNet-50, Evaluation Protocol=Linear probe, Reproduced=true2021.06 | 49.9 |