DreamVAR: Taming Reinforced Visual Autoregressive Model for High-Fidelity Subject-Driven Image Generation
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910005499789312 |
|---|---|
| author | Jiang, Xin Chen, Jingwen Li, Yehao Pan, Yingwei Chen, Kezhou Li, Zechao Yao, Ting Mei, Tao |
| author_facet | Jiang, Xin Chen, Jingwen Li, Yehao Pan, Yingwei Chen, Kezhou Li, Zechao Yao, Ting Mei, Tao |
| contents | Recent advances in subject-driven image generation using diffusion models have attracted considerable attention for their remarkable capabilities in producing high-quality images. Nevertheless, the potential of Visual Autoregressive (VAR) models, despite their unified architecture and efficient inference, remains underexplored. In this work, we present DreamVAR, a novel framework for subject-driven image synthesis built upon a VAR model that employs next-scale prediction. Technically, multi-scale features of the reference subject are first extracted by a visual tokenizer. Instead of interleaving these conditional features with target image tokens across scales, our DreamVAR pre-fills the full subject feature sequence prior to predicting target image tokens. This design simplifies autoregressive dependencies and mitigates the train-test discrepancy in multi-scale conditioning scenario within the VAR paradigm. DreamVAR further incorporates reinforcement learning to jointly enhance semantic alignment and subject consistency. Extensive experiments demonstrate that DreamVAR achieves superior appearance preservation compared to leading diffusion-based methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_22507 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DreamVAR: Taming Reinforced Visual Autoregressive Model for High-Fidelity Subject-Driven Image Generation Jiang, Xin Chen, Jingwen Li, Yehao Pan, Yingwei Chen, Kezhou Li, Zechao Yao, Ting Mei, Tao Computer Vision and Pattern Recognition Recent advances in subject-driven image generation using diffusion models have attracted considerable attention for their remarkable capabilities in producing high-quality images. Nevertheless, the potential of Visual Autoregressive (VAR) models, despite their unified architecture and efficient inference, remains underexplored. In this work, we present DreamVAR, a novel framework for subject-driven image synthesis built upon a VAR model that employs next-scale prediction. Technically, multi-scale features of the reference subject are first extracted by a visual tokenizer. Instead of interleaving these conditional features with target image tokens across scales, our DreamVAR pre-fills the full subject feature sequence prior to predicting target image tokens. This design simplifies autoregressive dependencies and mitigates the train-test discrepancy in multi-scale conditioning scenario within the VAR paradigm. DreamVAR further incorporates reinforcement learning to jointly enhance semantic alignment and subject consistency. Extensive experiments demonstrate that DreamVAR achieves superior appearance preservation compared to leading diffusion-based methods. |
| title | DreamVAR: Taming Reinforced Visual Autoregressive Model for High-Fidelity Subject-Driven Image Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2601.22507 |