Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2602.02033 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908807543652352 |
|---|---|
| author | Lu, Shuo Wang, Haohan Feng, Wei Wang, Weizhen Zhang, Shen Li, Yaoyu Ma, Ao Zhang, Zheng Lv, Jingjing Shen, Junjie Law, Ching Zhan, Bing Xu, Yuan Yao, Huizai Yu, Yongcan Si, Chenyang Liang, Jian |
| author_facet | Lu, Shuo Wang, Haohan Feng, Wei Wang, Weizhen Zhang, Shen Li, Yaoyu Ma, Ao Zhang, Zheng Lv, Jingjing Shen, Junjie Law, Ching Zhan, Bing Xu, Yuan Yao, Huizai Yu, Yongcan Si, Chenyang Liang, Jian |
| contents | Advertising image generation has increasingly focused on online metrics like Click-Through Rate (CTR), yet existing approaches adopt a ``one-size-fits-all" strategy that optimizes for overall CTR while neglecting preference diversity among user groups. This leads to suboptimal performance for specific groups, limiting targeted marketing effectiveness. To bridge this gap, we present \textit{One Size, Many Fits} (OSMF), a unified framework that aligns diverse group-wise click preferences in large-scale advertising image generation. OSMF begins with product-aware adaptive grouping, which dynamically organizes users based on their attributes and product characteristics, representing each group with rich collective preference features. Building on these groups, preference-conditioned image generation employs a Group-aware Multimodal Large Language Model (G-MLLM) to generate tailored images for each group. The G-MLLM is pre-trained to simultaneously comprehend group features and generate advertising images. Subsequently, we fine-tune the G-MLLM using our proposed Group-DPO for group-wise preference alignment, which effectively enhances each group's CTR on the generated images. To further advance this field, we introduce the Grouped Advertising Image Preference Dataset (GAIP), the first large-scale public dataset of group-wise image preferences, including around 600K groups built from 40M users. Extensive experiments demonstrate that our framework achieves the state-of-the-art performance in both offline and online settings. Our code and datasets will be released at https://github.com/JD-GenX/OSMF. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_02033 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | One Size, Many Fits: Aligning Diverse Group-Wise Click Preferences in Large-Scale Advertising Image Generation Lu, Shuo Wang, Haohan Feng, Wei Wang, Weizhen Zhang, Shen Li, Yaoyu Ma, Ao Zhang, Zheng Lv, Jingjing Shen, Junjie Law, Ching Zhan, Bing Xu, Yuan Yao, Huizai Yu, Yongcan Si, Chenyang Liang, Jian Computer Vision and Pattern Recognition Artificial Intelligence Multimedia Advertising image generation has increasingly focused on online metrics like Click-Through Rate (CTR), yet existing approaches adopt a ``one-size-fits-all" strategy that optimizes for overall CTR while neglecting preference diversity among user groups. This leads to suboptimal performance for specific groups, limiting targeted marketing effectiveness. To bridge this gap, we present \textit{One Size, Many Fits} (OSMF), a unified framework that aligns diverse group-wise click preferences in large-scale advertising image generation. OSMF begins with product-aware adaptive grouping, which dynamically organizes users based on their attributes and product characteristics, representing each group with rich collective preference features. Building on these groups, preference-conditioned image generation employs a Group-aware Multimodal Large Language Model (G-MLLM) to generate tailored images for each group. The G-MLLM is pre-trained to simultaneously comprehend group features and generate advertising images. Subsequently, we fine-tune the G-MLLM using our proposed Group-DPO for group-wise preference alignment, which effectively enhances each group's CTR on the generated images. To further advance this field, we introduce the Grouped Advertising Image Preference Dataset (GAIP), the first large-scale public dataset of group-wise image preferences, including around 600K groups built from 40M users. Extensive experiments demonstrate that our framework achieves the state-of-the-art performance in both offline and online settings. Our code and datasets will be released at https://github.com/JD-GenX/OSMF. |
| title | One Size, Many Fits: Aligning Diverse Group-Wise Click Preferences in Large-Scale Advertising Image Generation |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Multimedia |
| url | https://arxiv.org/abs/2602.02033 |