Qwen-Image Technical Report
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_ | 1866913974592733184 |
|---|---|
| author | Wu, Chenfei Li, Jiahao Zhou, Jingren Lin, Junyang Gao, Kaiyuan Yan, Kun Yin, Sheng-ming Bai, Shuai Xu, Xiao Chen, Yilei Chen, Yuxiang Tang, Zecheng Zhang, Zekai Wang, Zhengyi Yang, An Yu, Bowen Cheng, Chen Liu, Dayiheng Li, Deqing Zhang, Hang Meng, Hao Wei, Hu Ni, Jingyuan Chen, Kai Cao, Kuan Peng, Liang Qu, Lin Wu, Minggang Wang, Peng Yu, Shuting Wen, Tingkun Feng, Wensen Xu, Xiaoxiao Wang, Yi Zhang, Yichang Zhu, Yongqiang Wu, Yujia Cai, Yuxuan Liu, Zenan |
| author_facet | Wu, Chenfei Li, Jiahao Zhou, Jingren Lin, Junyang Gao, Kaiyuan Yan, Kun Yin, Sheng-ming Bai, Shuai Xu, Xiao Chen, Yilei Chen, Yuxiang Tang, Zecheng Zhang, Zekai Wang, Zhengyi Yang, An Yu, Bowen Cheng, Chen Liu, Dayiheng Li, Deqing Zhang, Hang Meng, Hao Wei, Hu Ni, Jingyuan Chen, Kai Cao, Kuan Peng, Liang Qu, Lin Wu, Minggang Wang, Peng Yu, Shuting Wen, Tingkun Feng, Wensen Xu, Xiaoxiao Wang, Yi Zhang, Yichang Zhu, Yongqiang Wu, Yujia Cai, Yuxuan Liu, Zenan |
| contents | We present Qwen-Image, an image generation foundation model in the Qwen series that achieves significant advances in complex text rendering and precise image editing. To address the challenges of complex text rendering, we design a comprehensive data pipeline that includes large-scale data collection, filtering, annotation, synthesis, and balancing. Moreover, we adopt a progressive training strategy that starts with non-text-to-text rendering, evolves from simple to complex textual inputs, and gradually scales up to paragraph-level descriptions. This curriculum learning approach substantially enhances the model's native text rendering capabilities. As a result, Qwen-Image not only performs exceptionally well in alphabetic languages such as English, but also achieves remarkable progress on more challenging logographic languages like Chinese. To enhance image editing consistency, we introduce an improved multi-task training paradigm that incorporates not only traditional text-to-image (T2I) and text-image-to-image (TI2I) tasks but also image-to-image (I2I) reconstruction, effectively aligning the latent representations between Qwen2.5-VL and MMDiT. Furthermore, we separately feed the original image into Qwen2.5-VL and the VAE encoder to obtain semantic and reconstructive representations, respectively. This dual-encoding mechanism enables the editing module to strike a balance between preserving semantic consistency and maintaining visual fidelity. Qwen-Image achieves state-of-the-art performance, demonstrating its strong capabilities in both image generation and editing across multiple benchmarks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_02324 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Qwen-Image Technical Report Wu, Chenfei Li, Jiahao Zhou, Jingren Lin, Junyang Gao, Kaiyuan Yan, Kun Yin, Sheng-ming Bai, Shuai Xu, Xiao Chen, Yilei Chen, Yuxiang Tang, Zecheng Zhang, Zekai Wang, Zhengyi Yang, An Yu, Bowen Cheng, Chen Liu, Dayiheng Li, Deqing Zhang, Hang Meng, Hao Wei, Hu Ni, Jingyuan Chen, Kai Cao, Kuan Peng, Liang Qu, Lin Wu, Minggang Wang, Peng Yu, Shuting Wen, Tingkun Feng, Wensen Xu, Xiaoxiao Wang, Yi Zhang, Yichang Zhu, Yongqiang Wu, Yujia Cai, Yuxuan Liu, Zenan Computer Vision and Pattern Recognition We present Qwen-Image, an image generation foundation model in the Qwen series that achieves significant advances in complex text rendering and precise image editing. To address the challenges of complex text rendering, we design a comprehensive data pipeline that includes large-scale data collection, filtering, annotation, synthesis, and balancing. Moreover, we adopt a progressive training strategy that starts with non-text-to-text rendering, evolves from simple to complex textual inputs, and gradually scales up to paragraph-level descriptions. This curriculum learning approach substantially enhances the model's native text rendering capabilities. As a result, Qwen-Image not only performs exceptionally well in alphabetic languages such as English, but also achieves remarkable progress on more challenging logographic languages like Chinese. To enhance image editing consistency, we introduce an improved multi-task training paradigm that incorporates not only traditional text-to-image (T2I) and text-image-to-image (TI2I) tasks but also image-to-image (I2I) reconstruction, effectively aligning the latent representations between Qwen2.5-VL and MMDiT. Furthermore, we separately feed the original image into Qwen2.5-VL and the VAE encoder to obtain semantic and reconstructive representations, respectively. This dual-encoding mechanism enables the editing module to strike a balance between preserving semantic consistency and maintaining visual fidelity. Qwen-Image achieves state-of-the-art performance, demonstrating its strong capabilities in both image generation and editing across multiple benchmarks. |
| title | Qwen-Image Technical Report |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2508.02324 |