Long-form Multimodal Storytelling
0.0674BERTScoreOurs
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| OursDescription=Full integrated design with Recursive Narrative Bank and speaker conditioning2025.05 | 0.0674 | 1.8726 | 27.4822 | 16.2412 | |
| BLIP2025.05 | 0.0674 | 1.8726 | 27.4822 | 17.0377 | |
| Speech w/o renderingRendering=False, Note=Ablation of speaker conditioning2025.05 | 0.0674 | 1.8726 | 27.4822 | 49.6302 | |
| Bark2025.05 | 0.0674 | 1.8726 | 27.4822 | 16.4668 | |
| Tacotron 22025.05 | 0.0674 | 1.8726 | 27.4822 | 16.4484 | |
| Without KeyFrameKeyFrame=False, Note=Ablation of visual grounding2025.05 | 0.0392 | 0.8609 | 25.3061 | 22.4406 | |
| VloggerNote=Integrated into dialogue and speech framework2025.05 | 0.0094 | 0.2648 | 26.1703 | 18.6999 | |
| Without RNBRNB=False, Note=Ablation of Recursive Narrative Bank2025.05 | 0.0094 | 0.5268 | 27.2259 | 16.5604 | |
| MochiNote=Integrated into dialogue and speech framework2025.05 | 0.0092 | 0.1272 | 26.6107 | 17.7431 |