Visual Narrative Generation on MSBench Multi-subject short-frame visual narratives
0.765CLIP-IMS-Diffusion + 1Prompt1Story
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MS-Diffusion + 1Prompt1StoryResolution=1024x1024, Sampling Steps=30, Hardware=RTX 40902026.02 | 0.765 | 0.314 | 0.491 | 7.88 | 21.25 | |
| Nano-BananaResolution=1024x1024, Sampling Steps=30, Hardware=RTX 40902026.02 | 0.737 | 0.391 | 0.195 | — | — | |
| FluxKontextResolution=1024x1024, Sampling Steps=30, Hardware=RTX 40902026.02 | 0.732 | 0.373 | 0.212 | 32.31 | 22.99 | |
| Qwen-EditResolution=1024x1024, Sampling Steps=30, Hardware=RTX 40902026.02 | 0.718 | 0.386 | 0.206 | — | — | |
| MS-DiffusionResolution=1024x1024, Sampling Steps=30, Hardware=RTX 40902026.02 | 0.687 | 0.318 | 0.323 | 5.36 | 19.47 | |
| StoryTailorResolution=1024x1024, Sampling Steps=30, Hardware=RTX 40902026.02 | 0.671 | 0.406 | 0.189 | 7.02 | 20.01 |