Scene Captioning on WHOOPS! RGBP seen scenes (test)
80.46CIDErQwen3-VL-8B-Instruct RGBP
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen3-VL-8B-Instruct RGBPBackbone=Qwen3-VL-8B, Dataset=RGBP2026.04 | 80.46 | 16.07 | |
| ProCapBackbone=TinyLlama-1.1B2026.04 | 69.95 | 21.87 | |
| ProCapBackbone=OpenLLaMA-3B2026.04 | 69.46 | 22.08 | |
| Qwen3-VL-8B-Instruct RGBPBackbone=Qwen3-VL-8B, Dataset=RGBP2026.04 | 37.7 | 23.59 | |
| ProCapBackbone=Vicuna-1.5-7B2026.04 | 36.53 | 22.7 | |
| ProCapBackbone=OPT-2.7B2026.04 | 28.19 | 24.11 | |
| ProCapBackbone=Vicuna-1.5-7B2026.04 | 22.33 | 6.02 | |
| ProCapBackbone=OpenLLaMA-3B2026.04 | 19.37 | 5.43 | |
| ProCapBackbone=OPT-2.7B2026.04 | 15.04 | 4.4 | |
| ProCapBackbone=TinyLlama-1.1B2026.04 | 11.88 | 3.62 | |
| Qwen3-VL-8B-InstructParameters=8B2026.04 | 11.19 | 10.97 | |
| Qwen3-VL-4B-InstructParameters=4B2026.04 | 10.35 | 10.91 | |
| Qwen3-VL-2B-InstructParameters=2B2026.04 | 9.82 | 11.05 | |
| FastVLM-7BParameters=7B2026.04 | 6.25 | 6.77 | |
| FastVLM-0.5BParameters=0.5B2026.04 | 5.54 | 5.85 | |
| FastVLM-1.5BParameters=1.5B2026.04 | 4.14 | 6.09 | |
| FastVLM-7BParameters=7B2026.04 | 2.21 | 11.98 | |
| Qwen3-VL-4B-InstructParameters=4B2026.04 | 2.02 | 12.15 | |
| Qwen3-VL-2B-InstructParameters=2B2026.04 | 2.01 | 12.83 | |
| Qwen3-VL-8B-InstructParameters=8B2026.04 | 1.97 | 13.07 | |
| FastVLM-1.5BParameters=1.5B2026.04 | 1.74 | 9.92 | |
| FastVLM-0.5BParameters=0.5B2026.04 | 1.59 | 10.17 |