EUVIMEDEUVIMED entdecken
Crossref

RFI: Rectified Flow Intervention for Mitigating Object Hallucination in Large Vision-Language Models

Junyu Cheng, Zhibiao Liang, Yidong Chen, Shuangyin Li

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (5) · 3255-3263

DOI: 10.1609/aaai.v40i5.37320 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Junyu Cheng und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Large Vision-Language Models (LVLMs) have demonstrated remarkable capabilities in multimodal understanding and generation by integrating visual and textual data. However, these models frequently exhibit object hallucination problems: generating outputs that are inconsistent with the input image. Existing improved methods for mitigating hallucinations still suffer from two key limitations: dynamic approaches based on logits or attention mechanisms risk suppressing valuable linguistic priors, whereas static methods that employ fixed intervention vectors lack the flexibility to adapt to diverse images and questions. To address these issues, we propose RFI (Rectified Flow Intervention), a novel approach that harnesses the linear trajectory design of rectified flow for input-specific adaptation and employs gradient correction to ensure coherent generation, effectively combining the adaptability of dynamic methods with the stability of static ones. RFI dynamically predicts latent-space intervention vectors while requiring only a single forward pass in LVLMs per question, achieving computational efficiency (1.09x latency overhead for 100 new tokens). Extensive experiments show RFI significantly reduces hallucinations, achieving superior performance compared to existing advanced methods, highlighting its effectiveness as a lightweight plug-and-play method for reducing LVLM's hallucination in practical applications.

Zitieren

APA

Cheng, J., Liang, Z., Chen, Y., & Li, S. (2026). RFI: Rectified Flow Intervention for Mitigating Object Hallucination in Large Vision-Language Models. Proceedings of the AAAI Conference on Artificial Intelligence, 40(5), 3255-3263. https://doi.org/10.1609/aaai.v40i5.37320

Vancouver

Cheng J, Liang Z, Chen Y, Li S. RFI: Rectified Flow Intervention for Mitigating Object Hallucination in Large Vision-Language Models. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(5):3255-3263. doi:10.1609/aaai.v40i5.37320

BibTeX
@article{Cheng2026,
  author = {Cheng, Junyu and Liang, Zhibiao and Chen, Yidong and Li, Shuangyin},
  title = {RFI: Rectified Flow Intervention for Mitigating Object Hallucination in Large Vision-Language Models},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {5},
  pages = {3255--3263},
  doi = {10.1609/aaai.v40i5.37320}
}
Kostenlos starten →