EUVIMEDEUVIMED entdecken
Crossref

SatireDecoder: Visual Cascaded Decoupling for Enhancing Satirical Image Comprehension

Yue Jiang, Haiwei Xue, Minghao Han, Mingcheng Li, Xiaolu Hou, Dingkang Yang, Lihua Zhang, Xu Zheng

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (7) · 5468-5476

DOI: 10.1609/aaai.v40i7.37464 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Yue Jiang und 7 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Satire, a form of artistic expression combining humor with implicit critique, holds significant social value by illuminating societal issues. Despite its cultural and societal significance, satire comprehension, particularly in purely visual forms, remains a challenging task for current vision-language models. This task requires not only detecting satire but also deciphering its nuanced meaning and identifying the implicated entities. Existing models often fail to effectively integrate local entity relationships with global context, leading to misinterpretation, comprehension biases, and hallucinations. To address these limitations, we propose SatireDecoder, a training-free framework designed to enhance satirical image comprehension. Our approach proposes a multi-agent system performing visual cascaded decoupling to decompose images into fine-grained local and global semantic representations. In addition, we introduce a chain-of-thought reasoning strategy guided by uncertainty analysis, which breaks down the complex satire comprehension process into sequential subtasks with minimized uncertainty. Our method significantly improves interpretive accuracy while reducing hallucinations. Experimental results validate that SatireDecoder outperforms existing baselines in comprehending visual satire, offering a promising direction for vision-language reasoning in nuanced, high-level semantic tasks.

Zitieren

APA

Jiang, Y., Xue, H., Han, M., Li, M., Hou, X., Yang, D., Zhang, L., & Zheng, X. (2026). SatireDecoder: Visual Cascaded Decoupling for Enhancing Satirical Image Comprehension. Proceedings of the AAAI Conference on Artificial Intelligence, 40(7), 5468-5476. https://doi.org/10.1609/aaai.v40i7.37464

Vancouver

Jiang Y, Xue H, Han M, Li M, Hou X, Yang D, et al. SatireDecoder: Visual Cascaded Decoupling for Enhancing Satirical Image Comprehension. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(7):5468-5476. doi:10.1609/aaai.v40i7.37464

BibTeX
@article{Jiang2026,
  author = {Jiang, Yue and Xue, Haiwei and Han, Minghao and Li, Mingcheng and Hou, Xiaolu and Yang, Dingkang and Zhang, Lihua and Zheng, Xu},
  title = {SatireDecoder: Visual Cascaded Decoupling for Enhancing Satirical Image Comprehension},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {7},
  pages = {5468--5476},
  doi = {10.1609/aaai.v40i7.37464}
}
Kostenlos starten →