论文

Mem2Fair:无需训练的文生视频公平性在线控制框架

Mem2Fair: Online Fairness Control for Text-to-Video Generation via Fair Memory

精选理由

做文生视频应用的人可以看看这篇:Mem2Fair 不用改模型、不用重新训练,靠一块按概念记录生成结果的记忆来动态调整提示词,就能纠偏'护士默认生成女性'这类偏见。跟训练类方法比它不动模型权重,跟采样类方法比它省掉部署前的批量生成,黑盒 API 模型也能直接套用。

arXiv 论文提出 Mem2Fair,针对文生视频(T2V)模型在提示词未指定受保护属性时产生的系统性偏见问题,例如描述 nurse 的提示词会不成比例地生成女性护士视频。

正文

该框架将属性级公平性建模为首个条件化在线可达性(online approachability)问题,核心机制是维护按概念划分的 Fair Memory:在推理阶段用已生成视频作为反馈,根据每个概念与目标分布的偏差,动态选择提示词层面的属性干预。

这一路径同时绕开了两类传统方案的缺陷:训练方法需要架构级访问且可能破坏时序一致性,采样方法则需要在部署前生成大量视频。实验显示 Mem2Fair 在多个人口属性、提示词设置和目标分布上持续优于未修改生成器与现有基线,提示词-视频对齐度和视频质量保持相当水平,并且在商业黑盒 T2V 模型上同样有效。

原文 · arXiv cs.AI

Mem2Fair: Online Fairness Control for Text-to-Video Generation via Fair Memory

Text-to-video (T2V) generation models can produce systematically unfair demographic outcomes when prompts leave protected attributes unspecified. For example, prompts describing nurses without specifying gender may disproportionately generate videos depicting female nurses. Mitigating unfairness in T2V generation is particularly challenging: training-based methods that modify model internals require architecture-specific access and risk disrupting temporal consistency, while sampling-based methods often incur costly pre-deployment video generation. We present the first conditional online approachability formulation of attribute-level fairness for T2V generation. Building on this formulation, we introduce Mem2Fair, a plug-and-play framework that maintains concept-conditioned Fair Memory. Using generated videos at inference time as feedback, Mem2Fair selects prompt-level attribute interventions based on the deviation from each concept's prescribed fairness target, without training or separate pre-deployment sampling. Experiments demonstrate that Mem2Fair consistently improves fairness over the unmodified generator and existing baselines across multiple demographic attributes, prompt settings, and target distributions, while maintaining comparable prompt-video alignment and video quality. We further demonstrate its effectiveness on commercial black-box T2V models.

产业坐标与信源核验

本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:Mem2Fair、文生视频、AI公平性、开源模型,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。

信源核验说明:本文原始线索采编自「arXiv cs.AI」(@Pengyu Chen, Jiaming Zhang, Xiaohua Feng, Yuyuan Li, Chaochao Chen),经 AITOP 多维评分管线结构化提炼。