X-Omni: Reinforcement Learning Makes Discrete Autoregressive Image Generative Models Great Again
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866911082041311232 |
|---|---|
| author | Geng, Zigang Wang, Yibing Ma, Yeyao Li, Chen Rao, Yongming Gu, Shuyang Zhong, Zhao Lu, Qinglin Hu, Han Zhang, Xiaosong Linus Wang, Di Jiang, Jie |
| author_facet | Geng, Zigang Wang, Yibing Ma, Yeyao Li, Chen Rao, Yongming Gu, Shuyang Zhong, Zhao Lu, Qinglin Hu, Han Zhang, Xiaosong Linus Wang, Di Jiang, Jie |
| contents | Numerous efforts have been made to extend the ``next token prediction'' paradigm to visual contents, aiming to create a unified approach for both image generation and understanding. Nevertheless, attempts to generate images through autoregressive modeling with discrete tokens have been plagued by issues such as low visual fidelity, distorted outputs, and failure to adhere to complex instructions when rendering intricate details. These shortcomings are likely attributed to cumulative errors during autoregressive inference or information loss incurred during the discretization process. Probably due to this challenge, recent research has increasingly shifted toward jointly training image generation with diffusion objectives and language generation with autoregressive objectives, moving away from unified modeling approaches. In this work, we demonstrate that reinforcement learning can effectively mitigate artifacts and largely enhance the generation quality of a discrete autoregressive modeling method, thereby enabling seamless integration of image and language generation. Our framework comprises a semantic image tokenizer, a unified autoregressive model for both language and images, and an offline diffusion decoder for image generation, termed X-Omni. X-Omni achieves state-of-the-art performance in image generation tasks using a 7B language model, producing images with high aesthetic quality while exhibiting strong capabilities in following instructions and rendering long texts. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_22058 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | X-Omni: Reinforcement Learning Makes Discrete Autoregressive Image Generative Models Great Again Geng, Zigang Wang, Yibing Ma, Yeyao Li, Chen Rao, Yongming Gu, Shuyang Zhong, Zhao Lu, Qinglin Hu, Han Zhang, Xiaosong Linus Wang, Di Jiang, Jie Computer Vision and Pattern Recognition Numerous efforts have been made to extend the ``next token prediction'' paradigm to visual contents, aiming to create a unified approach for both image generation and understanding. Nevertheless, attempts to generate images through autoregressive modeling with discrete tokens have been plagued by issues such as low visual fidelity, distorted outputs, and failure to adhere to complex instructions when rendering intricate details. These shortcomings are likely attributed to cumulative errors during autoregressive inference or information loss incurred during the discretization process. Probably due to this challenge, recent research has increasingly shifted toward jointly training image generation with diffusion objectives and language generation with autoregressive objectives, moving away from unified modeling approaches. In this work, we demonstrate that reinforcement learning can effectively mitigate artifacts and largely enhance the generation quality of a discrete autoregressive modeling method, thereby enabling seamless integration of image and language generation. Our framework comprises a semantic image tokenizer, a unified autoregressive model for both language and images, and an offline diffusion decoder for image generation, termed X-Omni. X-Omni achieves state-of-the-art performance in image generation tasks using a 7B language model, producing images with high aesthetic quality while exhibiting strong capabilities in following instructions and rendering long texts. |
| title | X-Omni: Reinforcement Learning Makes Discrete Autoregressive Image Generative Models Great Again |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2507.22058 |