BLM-Guard：基于推理和政策对齐奖励的可解释多模态广告审核
BLM-Guard: Explainable Multimodal Ad Moderation with Chain-of-Thought and Policy-Aligned Rewards
AI总结：
BLM-Guard通过融合链式推理和政策对齐奖励，实现多模态广告的可解释性审核，提升审核的准确性与鲁棒性。
摘要：
短视频平台如今承载着大量多模态广告，其欺骗性的视觉、语音和字幕需要更细致、基于政策的审核，而社区安全过滤器无法满足。我们提出了BLM-Guard，一个用于商业广告的内容审核框架，该框架融合了链式推理与基于规则的政策原则以及一个指导者引导的奖励。一个基于规则的ICoT数据合成管道通过生成结构化的场景描述、推理链和标签来启动训练，从而减少标注成本。强化学习随后利用复合奖励来平衡因果连贯性与政策合规性。多任务架构模型了内模操纵（例如，夸张的图像）和跨模态不匹配（例如，字幕语音漂移），提升了鲁棒性。在真实短视频广告上的实验表明，BLM-Guard在准确性、一致性和泛化能力方面均优于强大的基线模型。
英文摘要：
Short-video platforms now host vast multimodal ads whose deceptive visuals, speech and subtitles demand finer-grained, policy-driven moderation than community safety filters. We present BLM-Guard, a content-audit framework for commercial ads that fuses Chain-of-Thought reasoning with rule-based policy principles and a critic-guided reward. A rule-driven ICoT data-synthesis pipeline jump-starts training by generating structured scene descriptions, reasoning chains and labels, cutting annotation costs. Reinforcement learning then refines the model using a composite reward balancing causal coherence with policy adherence. A multitask architecture models intra-modal manipulations (e.g., exaggerated imagery) and cross-modal mismatches (e.g., subtitle-speech drift), boosting robustness. Experiments on real short-video ads show BLM-Guard surpasses strong baselines in accuracy, consistency and generalization.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
