Visual Question Answering on IrSD-VQA
82AccuracySpatialPIN
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SpatialPINBackbone=GPT-4o2024.03 | 82 | — | — | — | — | |
| GPT-4oBackbone=GPT-4o2024.03 | 64 | — | — | — | — | |
| SpatialVLMBackbone=PaLM 2-E2024.03 | 54 | — | — | — | — | |
| GPT-4o2024.03 | — | 30 | 10 | 4 | 2 | |
| GPT-4o + SpatialPINBackbone=GPT-4o2024.03 | — | 100 | 68 | 54 | 38 | |
| SpatialVLMBackbone=PaLM 2-E2024.03 | — | 78 | 26 | 12 | 4 |