Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.08850 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917327509913600 |
|---|---|
| author | Zhang, Guofeng Wang, Angtian Fang, Jacob Zhiyuan Jiang, Liming Yang, Haotian Yuille, Alan Ma, Chongyang |
| author_facet | Zhang, Guofeng Wang, Angtian Fang, Jacob Zhiyuan Jiang, Liming Yang, Haotian Yuille, Alan Ma, Chongyang |
| contents | Real-world videos naturally portray complex interactions among distinct physical objects, effectively forming dynamic compositions of visual elements. However, most current video generation models synthesize scenes holistically and therefore lack mechanisms for explicit compositional manipulation. To address this limitation, we propose HECTOR, a generative pipeline that enables fine-grained compositional control. In contrast to prior methods,HECTOR supports hybrid reference conditioning, allowing generation to be simultaneously guided by static images and/or dynamic videos. Moreover, users can explicitly specify the trajectory of each referenced element, precisely controlling its location, scale, and speed (see Figure1). This design allows the model to synthesize coherent videos that satisfy complex spatiotemporal constraints while preserving high-fidelity adherence to references. Extensive experiments demonstrate that HECTOR achieves superior visual quality, stronger reference preservation, and improved motion controllability compared with existing approaches. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_08850 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | HECTOR: Hybrid Editable Compositional Object References for Video Generation Zhang, Guofeng Wang, Angtian Fang, Jacob Zhiyuan Jiang, Liming Yang, Haotian Yuille, Alan Ma, Chongyang Computer Vision and Pattern Recognition Real-world videos naturally portray complex interactions among distinct physical objects, effectively forming dynamic compositions of visual elements. However, most current video generation models synthesize scenes holistically and therefore lack mechanisms for explicit compositional manipulation. To address this limitation, we propose HECTOR, a generative pipeline that enables fine-grained compositional control. In contrast to prior methods,HECTOR supports hybrid reference conditioning, allowing generation to be simultaneously guided by static images and/or dynamic videos. Moreover, users can explicitly specify the trajectory of each referenced element, precisely controlling its location, scale, and speed (see Figure1). This design allows the model to synthesize coherent videos that satisfy complex spatiotemporal constraints while preserving high-fidelity adherence to references. Extensive experiments demonstrate that HECTOR achieves superior visual quality, stronger reference preservation, and improved motion controllability compared with existing approaches. |
| title | HECTOR: Hybrid Editable Compositional Object References for Video Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.08850 |