Visual Search on V* bench (test)
87Attribute RateIVM-Enhanced GPT4-V
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| IVM-Enhanced GPT4-VModel Category=Specific Visual Search Models, Base Model=GPT4-V, Enhancement=IVM (Instruction-guided Visual Masking)2024.05 | 87 | 72.4 | 81.2 | |
| SEALModel Category=Specific Visual Search Models2024.05 | 74.8 | 76.3 | 75.4 | |
| GPT4-VModel Category=Commercial Chatbots2024.05 | 51.3 | 60.5 | 55 | |
| LLaVA-1.5Model Category=Open-Sourced LMMs2024.05 | 43.5 | 56.6 | 48.7 | |
| Gemini-ProModel Category=Commercial Chatbots2024.05 | 40.9 | 59.2 | 48.2 | |
| BardModel Category=Commercial Chatbots2024.05 | 31.3 | 46.1 | 37.2 | |
| MiniGPT-4Model Category=Open-Sourced LMMs2024.05 | 30.4 | 50 | 38.2 | |
| BLIP2Model Category=Open-Sourced LMMs2024.05 | 27 | 53.9 | 37.7 | |
| OtterModel Category=Open-Sourced LMMs2024.05 | 27 | 56.6 | 38.7 | |
| InstructBLIPModel Category=Open-Sourced LMMs2024.05 | 25.2 | 47.4 | 34 |