Scene Captioning on nocaps RGBP seen scenes (val)
102.67CIDErQwen3-VL-8B-Instruct RGBP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Qwen3-VL-8B-Instruct RGBPBackbone=Qwen3-VL-8B, Dataset=RGBP2026.04 | 102.67 | 13.19 | |
| ProCapBackbone=OpenLLaMA-3B2026.04 | 42.06 | 6.66 | |
| ProCapBackbone=Vicuna-1.5-7B2026.04 | 39.93 | 6.59 | |
| Qwen3-VL-8B-Instruct RGBPBackbone=Qwen3-VL-8B, Dataset=RGBP2026.04 | 37.48 | 23.49 | |
| ProCapBackbone=OpenLLaMA-3B2026.04 | 37.14 | 22.75 | |
| ProCapBackbone=Vicuna-1.5-7B2026.04 | 36.39 | 22.88 | |
| ProCapBackbone=TinyLlama-1.1B2026.04 | 35.21 | 24.21 | |
| ProCapBackbone=OPT-2.7B2026.04 | 30.58 | 5.17 | |
| ProCapBackbone=TinyLlama-1.1B2026.04 | 30.45 | 4.8 | |
| ProCapBackbone=OPT-2.7B2026.04 | 27.63 | 24.45 | |
| Qwen3-VL-8B-InstructParameters=8B2026.04 | 15.57 | 10.01 | |
| Qwen3-VL-4B-InstructParameters=4B2026.04 | 14.93 | 10.47 | |
| Qwen3-VL-2B-InstructParameters=2B2026.04 | 14.85 | 10.38 | |
| FastVLM-7BParameters=7B2026.04 | 7.01 | 6.12 | |
| FastVLM-0.5BParameters=0.5B2026.04 | 6.16 | 5.68 | |
| FastVLM-1.5BParameters=1.5B2026.04 | 5.72 | 5.88 | |
| FastVLM-7BParameters=7B2026.04 | 2.21 | 10.79 | |
| Qwen3-VL-4B-InstructParameters=4B2026.04 | 2.19 | 11.94 | |
| Qwen3-VL-2B-InstructParameters=2B2026.04 | 2.12 | 12.17 | |
| Qwen3-VL-8B-InstructParameters=8B2026.04 | 2.09 | 13.15 | |
| FastVLM-1.5BParameters=1.5B2026.04 | 1.73 | 9.16 | |
| FastVLM-0.5BParameters=0.5B2026.04 | 1.6 | 9.47 |