Multi-modal Sarcasm Detection on MMSD 2.0
87.38AccuracyGDCNet
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| GDCNetModality=Multi-Modal2026.01 | 87.38 | 83.39 | 89.51 | 86.34 | |
| TFCDModality=Multi-Modal2026.01 | 86.54 | 82.46 | 87.95 | 84.31 | |
| CofiParaModality=Multi-Modal2026.01 | 85.66 | 85.79 | 85.43 | 85.61 | |
| multi-view CLIPModality=Multi-modal, Backbone=clip-vit-base-patch322023.07 | 85.64 | 80.33 | 88.24 | 84.1 | |
| ADsModality=Multi-Modal2026.01 | 85.6 | 85.28 | 85.6 | 85.41 | |
| Multi-view CLIPModality=Multi-Modal2026.01 | 85.14 | 80.33 | 88.24 | 84.09 | |
| MoBAModality=Multi-Modal2026.01 | 85.01 | 80.46 | 87.67 | 83.64 | |
| DIPModality=Multi-Modal2026.01 | 84.63 | 84.17 | 85.2 | 84.68 | |
| ProCritProtocol=Fine-tuned, Perspective=Adaptive2026.05 | 84 | 76.1 | 91.5 | 83.1 | |
| ProCritProtocol=Fine-tuned, Perspective=Adaptive, Generation Mode=Draft-only2026.05 | 82.5 | 75.7 | 87.4 | 81.1 | |
| Fixed-role ReasonerProtocol=Fine-tuned, Perspective=Predefined2026.05 | 80.8 | 73.1 | 87.9 | 79.8 | |
| Plain CoT ReasonerProtocol=Fine-tuned, Perspective=None2026.05 | 80.3 | 72.3 | 88 | 79.4 | |
| Att-BERTModality=Multi-modal2023.07 | 80.03 | 76.28 | 77.82 | 77.04 | |
| CMGCNModality=Multi-modal, External Knowledge=false2023.07 | 79.83 | 75.82 | 78.01 | 76.9 | |
| InCrossMGsModality=Multi-Modal2026.01 | 79.83 | 75.82 | 78.01 | 76.9 | |
| RoBERTaModality=Text2023.07 | 79.66 | 76.74 | 75.7 | 76.21 | |
| BERTModality=Text-Only2026.01 | 76.52 | 74.48 | 73.09 | 73.78 | |
| HKEModality=Multi-modal, External Knowledge=false2023.07 | 76.5 | 73.48 | 71.07 | 72.25 | |
| HKEModality=Multi-Modal2026.01 | 76.5 | 73.48 | 71.07 | 72.25 | |
| SMSDModality=Text2023.07 | 73.56 | 68.45 | 71.55 | 69.97 | |
| SMSDModality=Text-Only2026.01 | 73.56 | 68.45 | 71.55 | 69.97 | |
| Bi-LSTMModality=Text2023.07 | 72.48 | 68.02 | 68.08 | 68.05 | |
| BiLSTMModality=Text-Only2026.01 | 72.48 | 68.02 | 68.08 | 68.05 | |
| ViTModality=Image2023.07 | 72.02 | 65.26 | 74.83 | 69.72 | |
| ViTModality=Image-Only2026.01 | 72.02 | 65.26 | 74.83 | 69.72 | |
| TextCNNModality=Text2023.07 | 71.61 | 64.62 | 75.22 | 69.52 | |
| ProCritProtocol=Zero-shot, Perspective=Adaptive2026.05 | 71.6 | 66.7 | 68.2 | 67.4 | |
| Automatic Prompt EngineerProtocol=Zero-shot, Perspective=None2026.05 | 71.4 | 72.9 | 53.4 | 61.7 | |
| Zero-shot CoTProtocol=Zero-shot, Perspective=None2026.05 | 71.1 | 71.9 | 53.9 | 61.6 | |
| S3AgentProtocol=Zero-shot, Perspective=Predefined2026.05 | 70.7 | 71.2 | 53.6 | 61.2 | |
| HFMModality=Multi-modal2023.07 | 70.57 | 64.84 | 69.05 | 66.88 | |
| Generated Knowledge PromptingProtocol=Zero-shot, Perspective=None2026.05 | 70.2 | 76.1 | 44.7 | 56.4 | |
| Plan-and-SolveProtocol=Zero-shot, Perspective=None2026.05 | 69 | 74.4 | 42.5 | 54.1 | |
| Commander-GPTProtocol=Zero-shot, Perspective=Predefined2026.05 | 65.8 | 58.8 | 68.6 | 63.3 | |
| ResNetModality=Image2023.07 | 65.5 | 61.17 | 54.39 | 57.58 | |
| ResNetModality=Image-Only2026.01 | 65.5 | 61.17 | 54.39 | 57.58 | |
| IRONICProtocol=Zero-shot, Perspective=Predefined2026.05 | 64.5 | 72.5 | 28.4 | 40.8 |