Object Hallucination Evaluation on POPE MSCOCO (val)
88.1F1 ScoreDeepSeek-VL-7B
Evaluation Results
| Method | Links | |
|---|---|---|
| DeepSeek-VL-7BSize=4B < Size < 8B2024.02 | 88.1 | |
| MM1-3B-ChatSize=Size < 4B2024.02 | 87.4 | |
| Bunny-8BSize=4B < Size < 8B2024.02 | 87.2 | |
| TinyLLaVA-Phi-2-SigLIP-3.1BSize=Size < 4B2024.02 | 87.2 | |
| Bunny-4BSize=Size < 4B2024.02 | 87.2 | |
| SPHINX-Intern2Size=4B < Size < 8B2024.02 | 86.9 | |
| Imp-v1.5-4B-Phi3Size=Size < 4B2024.02 | 86.9 | |
| LLaVA-v1.5-13B (LORA)Size=> 8B2024.02 | 86.7 | |
| Mipha-3BSize=Size < 4B2024.02 | 86.7 | |
| MM1-7B-ChatSize=4B < Size < 8B2024.02 | 86.6 | |
| LLaVA-NeXT-7BSize=4B < Size < 8B2024.02 | 86.5 | |
| LLaVA-v1.5-7B (LORA)Size=4B < Size < 8B2024.02 | 86.4 | |
| VILA1.5-13BSize=> 8B2024.02 | 86.3 | |
| MiniCPM-V 2.0Size=Size < 4B2024.02 | 86.3 | |
| LLaVA-NeXT-13BSize=> 8B2024.02 | 86.2 | |
| LVIS-INSTRUCT4V-13BSize=> 8B2024.02 | 86 | |
| VILA1.5-3BSize=Size < 4B2024.02 | 85.9 | |
| mPLUG-Owl2Size=4B < Size < 8B2024.02 | 85.8 | |
| MobileVLM-V2-3BSize=Size < 4B2024.02 | 84.7 | |
| InstructBLIP-Vicuna-13BSize=> 8B2024.02 | 83.7 | |
| GPT-4V-1106Size=> 8B2024.02 | 75.4 |