针对多模态问答中的幻觉问题，提出分阶段验证框架提升答案真实性。
Multi-Stage Verification-Centric Framework for Mitigating Hallucination in Multi-Modal RAG
- 设计多阶段验证流程，融合查询路由与双路径生成机制。
- 在多轮多模态任务中获第3名，显著降低幻觉率。
- 适合对事实准确性要求高的真实场景应用。
本文介绍团队CRUISE为应对KDD Cup 2025 Meta Comprehensive RAG Benchmark for Multi-modal, Multi-turn（CRAG-MM）挑战所开发的技术方案。该挑战旨在解决现代视觉语言模型（VLMs）在面对第一人称图像、长尾实体及复杂多跳问题时易产生幻觉的核心缺陷。这一问题在用户需高准确度事实查询的真实应用场景中尤为严重。为此，我们提出一种以事实准确性与真实性为核心、兼顾效率的多阶段框架，包含轻量级查询路由、查询感知的检索与摘要流程、双路径生成以及事后验证模块。该保守策略有效降低了竞赛评分中因幻觉导致的严重扣分。该方案在任务1中取得第3名，验证了优先保障答案可靠性的有效性。代码已开源：https://github.com/Breezelled/KDD-Cup-2025-Meta-CRAG-MM。
原文摘要 · Abstract (English)
This paper presents the technical solution developed by team CRUISE for the KDD Cup 2025 Meta Comprehensive RAG Benchmark for Multi-modal, Multi-turn (CRAG-MM) challenge. The challenge aims to address a critical limitation of modern Vision Language Models (VLMs): their propensity to hallucinate, especially when faced with egocentric imagery, long-tail entities, and complex, multi-hop questions. This issue is particularly problematic in real-world applications where users pose fact-seeking queries that demand high factual accuracy across diverse modalities. To tackle this, we propose a robust, multi-stage framework that prioritizes factual accuracy and truthfulness over completeness. Our solution integrates a lightweight query router for efficiency, a query-aware retrieval and summarization pipeline, a dual-pathways generation and a post-hoc verification. This conservative strategy is designed to minimize hallucinations, which incur a severe penalty in the competition's scoring metric. Our approach achieved 3rd place in Task 1, demonstrating the effectiveness of prioritizing answer reliability in complex multi-modal RAG systems. Our implementation is available at https://github.com/Breezelled/KDD-Cup-2025-Meta-CRAG-MM .
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。