titleCommerceGen-Studio: An End-to-End Framework for Commercial-Grade Main-Image Generation
authorsXiaoli Xu, Zhengtao Wu, Zhengze Xu, Tao Liu, Yongchao Du, Zuan Gao, Xingjian Wang, Hao Yan, Mengting Chen, Wei Li, Bo Zheng
date2026-10-14
venuearXiv (preprint, not peer reviewed)
idarXiv:2610.20603
affiliationalibaba
upvotes40 (Hugging Face Daily Papers, 2026-11-01)
linksabs pdf hf
abstractWe present CommerceGen-Studio, an end-to-end framework for generating commercial-grade main images for e-commerce Stock Keeping Units (SKUs). Our framework is organized along a three-stage pipeline of information extraction, semantic rewriting, and multi-condition generation: Prompt-Eval-Refine-Loop Agent (PERL-Agent) distills heterogeneous SKU metadata into concise selling-point text lists through an iterative closed loop of VLM and LLM roles scored by a reference-free rubric. Layout Thinker Optimization (LTO) aligns the rewriter with both caption and rendered-image quality via alternating domain-specific policy optimization. SKU-DiT, trained under Reinforce-Distill-Reinforce (RDR) progressive training, conditions jointly on the white-background product image, the trademark crop, and the structured layout to hold subject fidelity, trademark integrity, and character-level text accuracy together. The framework runs in production at a scale of millions of main images per day with a latency of a few seconds per image, which we believe makes it among the first systems of this kind to operate at commercial scale. We also release CommerceGen-Bench, a benchmark of hard cases collected from live traffic across all product categories. On it, our framework achieves the highest overall scores among the evaluated open-source baselines on both editing tracks, with and without prompt enhancement. It also matches much larger closed-source models, and beats several of them at Instruction Following and Product Integration - the two properties that determine whether an image can be published.

cite

@misc{xu2026commercegen,
  title={CommerceGen-Studio: An End-to-End Framework for Commercial-Grade Main-Image Generation},
  author={Xiaoli Xu and Zhengtao Wu and Zhengze Xu and Tao Liu and Yongchao Du and Zuan Gao and Xingjian Wang and Hao Yan and Mengting Chen and Wei Li and Bo Zheng},
  year={2026},
  eprint={2610.20603},
  archivePrefix={arXiv},
  url={https://arxiv.org/abs/2610.20603}
}