Semantic Consistency Evaluation on DSG
84.3Average Answering AccuracyVisualPrompter
Evaluation results
| Method | Links | ||
|---|---|---|---|
| VisualPrompterGenerative Model=Flux-dev2025.06 | 84.3 | 83 | |
| VisualPrompterGenerative Model=Janus-Pro2025.06 | 82.6 | 83 | |
| NeuroPromptsGenerative Model=Janus-Pro2025.06 | 81.7 | 74.6 | |
| BaselineGenerative Model=Janus-Pro2025.06 | 79.5 | 78.4 | |
| BaselineGenerative Model=Flux-dev2025.06 | 79.1 | 78.4 | |
| PromptistGenerative Model=Janus-Pro2025.06 | 78 | 76.2 | |
| VisualPrompterGenerative Model=SD 2.12025.06 | 77 | 83 | |
| PromptistGenerative Model=Flux-dev2025.06 | 76.9 | 76.2 | |
| NeuroPromptsGenerative Model=Flux-dev2025.06 | 75.5 | 74.6 | |
| BaselineGenerative Model=SD 2.12025.06 | 72.1 | 78.4 | |
| VisualPrompterGenerative Model=SD 1.52025.06 | 69.5 | 83 | |
| NeuroPromptsGenerative Model=SD 2.12025.06 | 68.7 | 74.6 | |
| PromptistGenerative Model=SD 2.12025.06 | 68.7 | 76.2 | |
| BaselineGenerative Model=SD 1.52025.06 | 67.5 | 78.4 | |
| PromptistGenerative Model=SD 1.52025.06 | 65.1 | 76.2 | |
| NeuroPromptsGenerative Model=SD 1.52025.06 | 58.4 | 74.6 | |
| BeautifulPromptGenerative Model=Janus-Pro2025.06 | 55.3 | 53.5 | |
| BeautifulPromptGenerative Model=Flux-dev2025.06 | 51.5 | 53.5 | |
| BeautifulPromptGenerative Model=SD 2.12025.06 | 49.6 | 53.5 | |
| BeautifulPromptGenerative Model=SD 1.52025.06 | 47.9 | 53.5 |