基于规划器-评论家智能体修正机制的画像感知可信食谱生成
Profile-Aware Trustworthy Recipe Generation with Planner-Critic Agentic Remediation
AI总结：
提出PCAR框架，通过规划器与安全评论家智能体的修正循环，在用户画像约束下生成可信食谱，兼顾安全性与实用性。
AI中文摘要：
从食物图像生成食谱对智能烹饪辅助具有实用价值，但传统的一次性生成往往忽视用户特定的安全需求，如过敏、饮食限制和制备约束。我们提出PCAR，一个用于可信食谱生成的规划器-评论家智能体修正框架。PCAR将食谱规划与安全验证分离：规划器智能体提取食材并根据用户画像生成食谱草稿，而安全评论家智能体审核每份草稿，并在检测到违规时提供结构化反馈以进行修正。这种修正循环使得不安全的食谱能够被修订，而不是直接返回或丢弃。我们在真实食物图像上使用100个基准用户画像，跨四个骨干模型（包括专有模型和本地服务的开源模型）评估PCAR。结果表明，PCAR在具备能力的骨干模型下实现了强大的安全性和生成性能，同时保持了实用的食谱质量。
英文摘要：
Recipe generation from food images has practical value for intelligent cooking assistance, but traditional one-pass generation often overlooks user-specific safety requirements such as allergies, dietary restrictions, and preparation constraints. We propose PCAR, a Planner-Critic Agentic Remediation framework for trustworthy recipe generation. PCAR separates recipe planning from safety verification: a Planner Agent extracts ingredients and generates recipe drafts conditioned on the user profile, while a Safety Critic Agent audits each draft and provides structured feedback for remediation when violations are detected. This remediation loop enables unsafe recipes to be revised rather than directly returned or discarded. We evaluate PCAR on real food images with 100 benchmark user profiles across four backbone models, including proprietary and locally served open-source models. Results show that PCAR achieves strong safety and generation performance with capable backbone models, while preserving practical recipe quality.
