arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2601.21796cs.CLcs.AI
收藏

KID: 基于知识注入的双头学习用于知识引导的有害迷因检测

KID: Knowledge-Injected Dual-Head Learning for Knowledge-Grounded Harmful Meme Detection

School of Economics and Management, Beijing University of Posts and Telecommunications(经济管理学院,北京邮电大学);College of Computing, Beijing Information Science and Technology University(计算机学院,北京信息科技大学)

Yaocong Li, Leihan Zhang, Le Zhang, Qiang Yan

AI总结:

KID通过知识注入和双头学习框架,提升有害迷因检测的准确性与泛化能力。

摘要:

互联网迷因已成为社交媒体上数字文化的主要传播载体。然而,它们对隐喻和社会文化背景的依赖也使其成为传播有害内容的微妙工具,给自动化内容审核带来了重大挑战。现有方法主要集中在模态内和模态间的信号分析,而对隐性毒性理解往往依赖于背景知识,而这些知识并不显式地出现在迷因本身中。为了解决这一挑战,我们提出了KID,一种用于知识引导的有害迷因检测的知识注入双头学习框架。KID采用标签约束的蒸馏范式,将复杂的迷因理解分解为结构化的推理链,这些链明确地将视觉证据、背景知识和分类标签联系起来。这些链通过在迷因特定上下文中将外部知识引入学习过程。此外,KID采用双头架构,联合优化语义生成和分类目标,使语言推理对齐的同时保持决策边界的稳定性。在五个涵盖英语、中文和低资源孟加拉语的多语言数据集上的广泛实验表明,KID在二元和多标签有害迷因检测任务上实现了SOTA性能,优于先前最佳方法的2.1%至19.7%。消融研究进一步确认了知识注入和双头联合学习的有效性,突显了它们对稳健和可推广的迷因理解的互补贡献。代码和数据可在https://github.com/PotatoDog1669/KID上获得。

英文摘要:

Internet memes have become pervasive carriers of digital culture on social platforms. However, their heavy reliance on metaphors and sociocultural context also makes them subtle vehicles for harmful content, posing significant challenges for automated content moderation. Existing approaches primarily focus on intra-modal and inter-modal signal analysis, while the understanding of implicit toxicity often depends on background knowledge that is not explicitly present in the meme itself. To address this challenge, we propose KID, a Knowledge-Injected Dual-Head Learning framework for knowledge-grounded harmful meme detection. KID adopts a label-constrained distillation paradigm to decompose complex meme understanding into structured reasoning chains that explicitly link visual evidence, background knowledge, and classification labels. These chains guide the learning process by grounding external knowledge in meme-specific contexts. In addition, KID employs a dual-head architecture that jointly optimizes semantic generation and classification objectives, enabling aligned linguistic reasoning while maintaining stable decision boundaries. Extensive experiments on five multilingual datasets spanning English, Chinese, and low-resource Bengali demonstrate that KID achieves SOTA performance on both binary and multi-label harmful meme detection tasks, improving over previous best methods by 2.1%--19.7% across primary evaluation metrics. Ablation studies further confirm the effectiveness of knowledge injection and dual-head joint learning, highlighting their complementary contributions to robust and generalizable meme understanding. The code and data are available at https://github.com/PotatoDog1669/KID.

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。

↑