Image Captioning on MC-LLaVA
83.6Caption Recall (Single)GPT4o+Prompt
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GPT4o+PromptT (Tokens)=10^12024.11 | 83.6 | 81.6 | 83 | |
| MC-LLaVAT (Tokens)=10^1, Auxiliary Loss=true2024.11 | 77.2 | 78.4 | 77.5 | |
| MC-LLaVAT (Tokens)=10^1, Auxiliary Loss=false2024.11 | 74.1 | 76.3 | 75.4 | |
| PeKitT (Tokens)=02024.11 | 72.9 | 74.3 | 73.3 | |
| MyVLMT (Tokens)=12024.11 | 71.4 | — | 71.4 | |
| RAP-MLLMT (Tokens)=10^22024.11 | 71.1 | 74.8 | 72.3 | |
| Yo'LLaVA-ST (Tokens)=10^12024.11 | 70.1 | — | 70.1 | |
| Yo'LLaVA-MT (Tokens)=10^12024.11 | 62.2 | 61.1 | 61.9 | |
| LLaVA+PromptT (Tokens)=10^12024.11 | 10.8 | 16 | 12.3 | |
| LLaVAT (Tokens)=02024.11 | 9.6 | 5 | 8.2 | |
| LLaVA+PromptT (Tokens)=10^22024.11 | 5.4 | 12.2 | 7.5 |