让聊天助手更懂图像，智能推荐下一步该改什么。
What to Edit Next: Visually Aligned Image-Editing Follow-Up Suggestions in Conversational Systems
- 基于真实对话数据构建多轮编辑意图库，训练多模态推荐策略。
- 通过用户点击反馈优化策略，提升建议与实际选择的一致性。
- 引入视觉验证器减少修改建议与图像的视觉冲突，适合图像创作场景。
对话式助手正越来越多地推荐后续编辑以帮助用户持续完成任务。现有系统主要面向纯文本交互，图像生成类对话仍待探索。在图像生成任务中，有效的后续编辑建议需反映用户偏好、提供多样化方向，并可在当前图像上执行。我们从Qwen App收集了10万条真实的多轮图像生成对话样本，发现80.1%为图像依赖型，凸显多模态推荐的重要性。为此提出三阶段框架：第一阶段利用真实在线数据构建人工审核的合适后续编辑意图表，生成SFT目标并微调多模态策略；第二阶段通过用户点击反馈，采用多目标强化学习优化策略，对齐规则指导的SFT建议与实际用户选择；第三阶段引入视觉验证器作为额外训练监督，降低建议编辑与当前图像间的视觉不一致。大量实验表明，本框架在自动与人工评估中均显著优于基线。在百万级用户的线上随机A/B测试中，视觉不一致率从3.7%降至0.9%，推荐点击率提升32.70%，图像采纳率提升16.32%，用户平均对话轮次增加39.90%（均p<0.05）。项目页面：https://what-to-edit-next.github.io/
原文摘要 · Abstract (English)
Conversational assistants increasingly recommend follow-up edits to help users continue a task. Existing systems primarily target text-only interactions, leaving image-creation conversations underexplored. In image-creation tasks, useful follow-up edit suggestions must reflect user preferences, offer diverse directions, and remain executable on the current image. We collected 100,000 real multi-turn image-creation conversation samples from Qwen App and found that 80.1% are image-dependent, underscoring the need for multimodal recommendation. We address this setting with a three-stage framework. In Stage 1, we use real online data to build a human-reviewed table of appropriate follow-up editing intents, then create SFT targets and fine-tune a multimodal policy. In Stage 2, to align rule-guided SFT suggestions with actual user choices, we use user click feedback to optimize the policy through multi-objective reinforcement learning. In Stage 3, to reduce visual inconsistencies between suggested edits and the current image, we introduce a visual verifier as additional training supervision. Extensive experiments demonstrate that our framework significantly outperforms baselines on both automatic and human evaluations. In a live user-randomized A/B test with millions of users, our final framework reduces visual inconsistency from 3.7% to 0.9%. Furthermore, it significantly improves recommendation CTR by 32.70%, image take-away rate by 16.32%, and average conversation turns per user by 39.90% (all p<0.05). Project page: https://what-to-edit-next.github.io/
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。