Discriminative Task on AMBER Discrimination 1.0 (test)
76.7AccuracyOctopus
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| OctopusBackbone=LLaVA-1.5-7B, Method=Octopus2025.03 | 76.7 | 82.7 | |
| OctopusBackbone=InstructBLIP, Method=Octopus2025.03 | 74 | 79.7 | |
| AVISCBackbone=InstructBLIP, Method=AVISC2025.03 | 72.6 | 78.6 | |
| AVISCBackbone=LLaVA-1.5-7B, Method=AVISC2025.03 | 70.7 | 75.45 | |
| VCDBackbone=InstructBLIP, Method=VCD2025.03 | 69.65 | 75.9 | |
| M3IDBackbone=InstructBLIP, Method=M3ID2025.03 | 69.05 | 75.25 | |
| InstructBLIPBackbone=InstructBLIP, Method=Base2025.03 | 68.2 | 74.6 | |
| VCDBackbone=LLaVA-1.5-7B, Method=VCD2025.03 | 67.3 | 71.1 | |
| M3IDBackbone=LLaVA-1.5-7B, Method=M3ID2025.03 | 67.25 | 70.9 | |
| LLaVA-1.5-7BBackbone=LLaVA-1.5-7B, Method=Base2025.03 | 67 | 71.1 |