Object Counting Ranking on COCO-REM 4-imgs
0.871Kendall's TauQwen2.5-VL-7B + RARL
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen2.5-VL-7B + RARLBackbone=Qwen2.5-VL-7B, Training Strategy=RARL2026.01 | 0.871 | |
| Qwen2.5-VL-3B + RARLBackbone=Qwen2.5-VL-3B, Training Strategy=RARL2026.01 | 0.824 | |
| Qwen2.5-VL-3B + SFTBackbone=Qwen2.5-VL-3B, Training Strategy=SFT2026.01 | 0.774 | |
| Qwen2.5-VL-7B + SFTBackbone=Qwen2.5-VL-7B, Training Strategy=SFT2026.01 | 0.772 | |
| Qwen2.5-VL-3BBackbone=Qwen2.5-VL-3B, Training Strategy=Base2026.01 | 0.753 | |
| Qwen2.5-VL-7BBackbone=Qwen2.5-VL-7B, Training Strategy=Base2026.01 | 0.732 |