EUVIMEDEUVIMED entdecken
Crossref

Using Unreliable Pseudo-Labels for Label-Efficient Semantic Segmentation

Haochen Wang, Yuchao Wang, Yujun Shen, Junsong Fan, Yuxi Wang, Zhaoxiang Zhang

International Journal of Computer Vision · 2025-04 · 133 (4) · 1476-1498

DOI: 10.1007/s11263-024-02229-x ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Computer Vision, online erschienen am 08.10.2024. Verfasst von Haochen Wang und 5 weiteren Autor:innen. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 137 Literaturangaben. Gefördert von: National Natural Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The crux of label-efficient semantic segmentation is to produce high-quality pseudo-labels to leverage a large amount of unlabeled or weakly labeled data. A common practice is to select the highly confident predictions as the pseudo-ground-truths for each pixel, but it leads to a problem that most pixels may be left unused due to their unreliability. However, we argue that every pixel matters to the model training, even those unreliable and ambiguous pixels. Intuitively, an unreliable prediction may get confused among the top classes, however, it should be confident about the pixel not belonging to the remaining classes. Hence, such a pixel can be convincingly treated as a negative key to those most unlikely categories. Therefore, we develop an effective pipeline to make sufficient use of unlabeled data. Concretely, we separate reliable and unreliable pixels via the entropy of predictions, push each unreliable pixel to a category-wise queue that consists of negative keys, and manage to train the model with all candidate pixels. Considering the training evolution, we adaptively adjust the threshold for the reliable-unreliable partition. Experimental results on various benchmarks and training settings demonstrate the superiority of our approach over the state-of-the-art alternatives.

Förderung

  • National Natural Foundation of China

Zitieren

APA

Wang, H., Wang, Y., Shen, Y., Fan, J., Wang, Y., & Zhang, Z. (2025). Using Unreliable Pseudo-Labels for Label-Efficient Semantic Segmentation. International Journal of Computer Vision, 133(4), 1476-1498. https://doi.org/10.1007/s11263-024-02229-x

Vancouver

Wang H, Wang Y, Shen Y, Fan J, Wang Y, Zhang Z. Using Unreliable Pseudo-Labels for Label-Efficient Semantic Segmentation. International Journal of Computer Vision. 2025;133(4):1476-1498. doi:10.1007/s11263-024-02229-x

BibTeX
@article{Wang2025,
  author = {Wang, Haochen and Wang, Yuchao and Shen, Yujun and Fan, Junsong and Wang, Yuxi and Zhang, Zhaoxiang},
  title = {Using Unreliable Pseudo-Labels for Label-Efficient Semantic Segmentation},
  journal = {International Journal of Computer Vision},
  year = {2025},
  volume = {133},
  number = {4},
  pages = {1476--1498},
  doi = {10.1007/s11263-024-02229-x}
}
Kostenlos starten →