Identification on CFMS
83.25AccuracyLoRA FT
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| LoRA FTModel=InternVL2.5-8B2026.03 | 83.25 | 79.7 | 83.96 | 81.77 | |
| LoRA FTModel=Qwen2.5-VL-7B-Instruct2026.03 | 82.3 | 81.22 | 78.61 | 79.89 | |
| LoRA FTModel=LLaVA-1.5-7B2026.03 | 80.95 | 84.71 | 70.37 | 76.88 | |
| Qwen2.5-VL-7B-InstructPrompting Strategy=PGDS2026.03 | 78.01 | 76.92 | 75.76 | 76.34 | |
| GPT-4oPrompting Strategy=RAG 1-shot2026.03 | 77.62 | 68.2 | 94.18 | 79.1 | |
| Qwen2.5-VL-7B-InstructPrompting Strategy=RAG 1-shot2026.03 | 76.88 | 71.08 | 81.94 | 76.13 | |
| GPT-4oPrompting Strategy=Random 1-shot2026.03 | 76.36 | 70.08 | 86.41 | 77.39 | |
| Gemini-2.5-FlashPrompting Strategy=Random 1-shot2026.03 | 75.71 | 67.61 | 88.36 | 76.61 | |
| Gemini-2.5-FlashPrompting Strategy=RAG 1-shot2026.03 | 75 | 65.79 | 92.59 | 76.92 | |
| InternVL2.5-8BPrompting Strategy=PGDS2026.03 | 74.76 | 66.27 | 89.42 | 76.13 | |
| Qwen2.5-VL-7B-InstructPrompting Strategy=Random 1-shot2026.03 | 71.25 | 64.54 | 80.16 | 71.5 | |
| Gemini-2.5-FlashPrompting Strategy=Zero-shot2026.03 | 71.19 | 60.31 | 95.24 | 74.84 | |
| InternVL2.5-8BPrompting Strategy=RAG 1-shot2026.03 | 71.05 | 62.11 | 96.15 | 75.47 | |
| Qwen2.5-VL-7B-InstructPrompting Strategy=Zero-shot2026.03 | 70.71 | 61.64 | 73.54 | 69.33 | |
| GPT-4oPrompting Strategy=Zero-shot2026.03 | 70.24 | 60.67 | 96.3 | 74.44 | |
| InternVL2.5-8BPrompting Strategy=Random 1-shot2026.03 | 69.29 | 60.1 | 94.44 | 73.46 | |
| InternVL2.5-8BPrompting Strategy=Zero-shot2026.03 | 68.48 | 59.67 | 93.72 | 72.91 |