EUVIMEDEUVIMED entdecken
Crossref

Confidence-Aware Pseudo-Label Self-Correction for Weakly Supervised Visual Grounding

Yang Liu, Jiahua Zhang, Yue Wu, Zijing Zhao, Qingchao Chen, Yuxin Peng

IEEE Transactions on Pattern Analysis and Machine Intelligence · 2026-08 · 48 (8) · 8901-8918

DOI: 10.1109/tpami.2026.3673741 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026-08. Verfasst von Yang Liu und 5 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 79 Literaturangaben. Gefördert von: Beijing Nova Program, National Natural Science Foundation of China, Beijing Natural Science Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Weakly supervised visual grounding aims to locate a region in an image based on an input query sentence, without access to the mapping between image regions and queries during training. Current methods treat spatial grounding as an object retrieval task, relying on cross-modal similarity scores for proposal selection. However, they fail to address model overfitting caused by unreliable cross-modal similarity scores. To overcome this, we first propose the Confidence-aware Pseudo-label Learning (CPL) framework. CPL first generates diverse pseudo queries for region proposals, and then establishes reliable associations for model training based on the uni-modal similarity score. Secondly, we propose a cross-modal verification module based on the pretrained vision-language model to verify associations. However, the verification module is isolated from the grounding model, so it can only assess associations in a static manner, but not correct the suspicious ones. Finally, we introduce CPL++ to make two-fold improvements. For one thing, we upgrade the verification process based on the model's grounding loss value to identify suspicious associations dynamically and selectively leverage them in the training. For another, we propose a self-supervised association correction module to rectify suspicious associations, thereby mitigating the risk of error propagation. Experimental results on five datasets demonstrate the superiority of our approach.

Förderung

  • Beijing Nova Program, National Natural Science Foundation of China
  • Beijing Natural Science Foundation

Zitieren

APA

Liu, Y., Zhang, J., Wu, Y., Zhao, Z., Chen, Q., & Peng, Y. (2026). Confidence-Aware Pseudo-Label Self-Correction for Weakly Supervised Visual Grounding. IEEE Transactions on Pattern Analysis and Machine Intelligence, 48(8), 8901-8918. https://doi.org/10.1109/tpami.2026.3673741

Vancouver

Liu Y, Zhang J, Wu Y, Zhao Z, Chen Q, Peng Y. Confidence-Aware Pseudo-Label Self-Correction for Weakly Supervised Visual Grounding. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2026;48(8):8901-8918. doi:10.1109/tpami.2026.3673741

BibTeX
@article{Liu2026,
  author = {Liu, Yang and Zhang, Jiahua and Wu, Yue and Zhao, Zijing and Chen, Qingchao and Peng, Yuxin},
  title = {Confidence-Aware Pseudo-Label Self-Correction for Weakly Supervised Visual Grounding},
  journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
  year = {2026},
  volume = {48},
  number = {8},
  pages = {8901--8918},
  doi = {10.1109/tpami.2026.3673741}
}
Kostenlos starten →