Hateful Meme Classification On Harmeme
评估指标
AUROC
Accuracy
评测结果
各个模型在此基准测试上的表现结果
模型名称 | AUROC | Accuracy | Paper Title | Repository |
|---|---|---|---|---|
| ISSUES | 92.83 | 81.64 | Mapping Memes to Words for Multimodal Hateful Meme Classification | |
| Hate-CLIPper | 91.87 | 83.90 | Hate-CLIPper: Multimodal Hateful Meme Classification based on Cross-modal Interaction of CLIP Features | |
| LMM-RGCL (Qwen2VL-7B) | 93.2 | 88.1 | Improved Fine-Tuning of Large Multimodal Models for Hateful Meme Detection | |
| LMM-RGCL (Qwen2VL-2B) | 92.9 | 87.7 | Improved Fine-Tuning of Large Multimodal Models for Hateful Meme Detection | |
| RGCL | 91.80 | 87.00 | Improving Hateful Meme Detection through Retrieval-Guided Contrastive Learning | |
| PromptHate | 90.96 | 84.47 | Prompting for Multimodal Hateful Meme Classification | - |
| DisMultiHate | 86.39 | 81.24 | Disentangling Hate in Online Memes | - |
0 of 7 row(s) selected.