EUVIMEDEUVIMED entdecken
Crossref

CaSS: Category-Aware Semantic Segmentation With Vision–Language Priors

Quan Tang, Dengke Zhang, Xuhao Tang, Bin Wang, Cuifeng Du, Jun Jiang

IEEE Signal Processing Letters · 2026 · 33 · 1651-1655

DOI: 10.1109/lsp.2026.3684447 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Signal Processing Letters, 2026. Verfasst von Quan Tang und 5 weiteren Autor:innen. Die Arbeit nennt 30 Literaturangaben. Gefördert von: Major Key Project of PCL, China Postdoctoral Science Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Semantic image segmentation typically relies on static pixel-wise classifiers that operate over a fixed category space, making them insensitive to the actual semantic composition of a real-world image. In this work, we propose CaSS, a category-aware semantic segmentation framework that introduces image-level semantic priors to dynamically adapt pixel-level classification. Specifically, a pre-trained vision–language model is employed to infer the set of semantic categories present in an image, which is encoded as a structured category prior. This prior is then used to drive a lightweight dynamic parsing network that generates image-conditioned classifier parameters for pixel-wise segmentation. By explicitly constraining the classifier with category-aware priors, CaSS reduces interference from absent classes and enhances both intra-class consistency and inter-class discriminability. The proposed approach follows a large–small model collaboration paradigm, leveraging the strong semantic understanding of vision–language models while preserving efficient pixel-level inference. Extensive experiments on standard semantic segmentation benchmarks demonstrate that CaSS consistently improves segmentation accuracy over state-of-the-art methods with minimal parameter overhead.

Förderung

  • Major Key Project of PCL
  • China Postdoctoral Science Foundation

Zitieren

APA

Tang, Q., Zhang, D., Tang, X., Wang, B., Du, C., & Jiang, J. (2026). CaSS: Category-Aware Semantic Segmentation With Vision–Language Priors. IEEE Signal Processing Letters, 33, 1651-1655. https://doi.org/10.1109/lsp.2026.3684447

Vancouver

Tang Q, Zhang D, Tang X, Wang B, Du C, Jiang J. CaSS: Category-Aware Semantic Segmentation With Vision–Language Priors. IEEE Signal Processing Letters. 2026;33:1651-1655. doi:10.1109/lsp.2026.3684447

BibTeX
@article{Tang2026,
  author = {Tang, Quan and Zhang, Dengke and Tang, Xuhao and Wang, Bin and Du, Cuifeng and Jiang, Jun},
  title = {CaSS: Category-Aware Semantic Segmentation With Vision–Language Priors},
  journal = {IEEE Signal Processing Letters},
  year = {2026},
  volume = {33},
  pages = {1651--1655},
  doi = {10.1109/lsp.2026.3684447}
}
Kostenlos starten →