First Frame Is the Place to Go for Video Content Customization
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917357474021376 |
|---|---|
| author | Chen, Jingxi Li, Zongxia Liu, Zhichao Shi, Guangyao Wu, Xiyang Liu, Fuxiao Fermuller, Cornelia Feng, Brandon Y. Aloimonos, Yiannis |
| author_facet | Chen, Jingxi Li, Zongxia Liu, Zhichao Shi, Guangyao Wu, Xiyang Liu, Fuxiao Fermuller, Cornelia Feng, Brandon Y. Aloimonos, Yiannis |
| contents | What role does the first frame play in video generation models? Traditionally, it's viewed as the spatial-temporal starting point of a video, merely a seed for subsequent animation. In this work, we reveal a fundamentally different perspective: video models implicitly treat the first frame as a conceptual memory buffer that stores visual entities for later reuse during generation. Leveraging this insight, we show that it's possible to achieve robust and generalized video content customization in diverse scenarios, using only 20-50 training examples without architectural changes or large-scale finetuning. This unveils a powerful, overlooked capability of video generation models for reference-based video customization. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_15700 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | First Frame Is the Place to Go for Video Content Customization Chen, Jingxi Li, Zongxia Liu, Zhichao Shi, Guangyao Wu, Xiyang Liu, Fuxiao Fermuller, Cornelia Feng, Brandon Y. Aloimonos, Yiannis Computer Vision and Pattern Recognition What role does the first frame play in video generation models? Traditionally, it's viewed as the spatial-temporal starting point of a video, merely a seed for subsequent animation. In this work, we reveal a fundamentally different perspective: video models implicitly treat the first frame as a conceptual memory buffer that stores visual entities for later reuse during generation. Leveraging this insight, we show that it's possible to achieve robust and generalized video content customization in diverse scenarios, using only 20-50 training examples without architectural changes or large-scale finetuning. This unveils a powerful, overlooked capability of video generation models for reference-based video customization. |
| title | First Frame Is the Place to Go for Video Content Customization |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2511.15700 |