Visual Question Answering (VQA) 벤치마크
Visual Question Answering (VQA) on IllusionVQA
Accuracy
- 2024-03-23 — GPT4-Vision 4-shot: Accuracy 62.99
| Rank | Model | Accuracy | Paper | Code | Year |
|---|---|---|---|---|---|
| 1 | GPT4-Vision 4-shot | 62.99 | IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models | csebuetnlp/illusionvqa | 2024 |
| 2 | GPT4-Vision | 58.85 | IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models | csebuetnlp/illusionvqa | 2024 |
| 3 | Gemini-Pro 4-shot | 52.87 | IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models | csebuetnlp/illusionvqa | 2024 |
| 4 | Gemini-Pro | 51.26 | IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models | csebuetnlp/illusionvqa | 2024 |
| 5 | LLaVA-1.5-13B | 40 | IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models | csebuetnlp/illusionvqa | 2024 |
| 6 | CogVLM | 38.16 | IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models | csebuetnlp/illusionvqa | 2024 |
| 7 | InstructBLIP-13B | 34.25 | IllusionVQA: A Challenging Optical Illusion Dataset for Vision Language Models | csebuetnlp/illusionvqa | 2024 |