Visual Question Answering on DeepEval
64.3ACCOurs
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| OursBackbone=Qwen2.5-VL-3B2026.01 | 64.3 | 85.9 | 12.9 | |
| CCoTVenue=CVPR'24, Backbone=Qwen2.5-VL-3B2026.01 | 51.3 | 60.4 | 19.6 | |
| ToTVenue=NIPS'23, Backbone=Qwen2.5-VL-3B2026.01 | 50.8 | 79 | 22.4 | |
| GoTVenue=AAAI'24, Backbone=Qwen2.5-VL-3B2026.01 | 49.8 | 62.8 | 13.5 | |
| DDCoTVenue=NIPS'23, Backbone=Qwen2.5-VL-3B2026.01 | 49.2 | 62.4 | 18 | |
| CoTVenue=NIPS'22, Backbone=Qwen2.5-VL-3B2026.01 | 48.7 | 70.4 | 14.5 | |
| CoT-SCVenue=ICLR'23, Backbone=Qwen2.5-VL-3B2026.01 | 45.2 | 61.9 | 21.7 | |
| LADVenue=arXiv'25, Backbone=Qwen2.5-VL-3B2026.01 | 42.7 | 15.4 | 13.1 |