QANet: Query-Aware Multi-Modal Prior Refinement for Few-Shot Segmentation
IEEE Transactions on Circuits and Systems for Video Technology · 2026-07 · 36 (7) · 10721-10736
DOI: 10.1109/tcsvt.2026.3672454 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Circuits and Systems for Video Technology, 2026-07. Verfasst von Qiang Jiao und 3 weiteren Autor:innen. Die Arbeit nennt 52 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Fundamental Research Funds for the Central Universities, Major Scientific and Technological Innovation Project of Xianyang.
Abstract (Originalsprache)
Recent Few-Shot Segmentation (FSS) approaches incorporate vision-language models to improve segmentation by constructing multi-modal priors, including visual, CAM and textual priors. However, these priors are often suboptimal, suffering from background interference, incomplete activation and semantic misalignment. To address these limitations, we propose a query-aware multi-modal prior refinement framework (QANet) for FSS, in which the multi-modal priors are jointly refined for more accurate segmentation by fully exploiting target-relevant cues from query features. Specifically, QANet comprises two modules, i.e., a Query-aware Multi-modal Prior Refinement (QMPR) module for visual and CAM prior refinement, and a Query-Aware Textual Embedding and Prior Refinement (QTEPR) module for textual prior refinement. More specifically, in QMPR, a query self-attention bilateral-guided enhancement strategy and a query self-similarity prior refinement strategy are carefully designed for suppressing background interferences and recovering foreground completeness, respectively. And in QTEPR, such refined visual and CAM priors are first leveraged to extract target-relevant cues for visual-textual domain gap mitigation and text embedding enhancement, producing query-aware textual priors. These textual priors are then enhanced under the guidance of those refined multi-modal priors. Extensive experiments on PASCAL- $\boldsymbol {5^{i}}$ and COCO- $\boldsymbol {20^{i}}$ datasets demonstrate that QANet achieves new state-of-the-art performance. Moreover, QANet remains competitive on cross-domain FSS and weak-label FSS tasks.
Förderung
- National Natural Science Foundation of China
- Fundamental Research Funds for the Central Universities
- Major Scientific and Technological Innovation Project of Xianyang
Zitieren
APA
Jiao, Q., Shi, M., Yang, Y., & Zhang, Q. (2026). QANet: Query-Aware Multi-Modal Prior Refinement for Few-Shot Segmentation. IEEE Transactions on Circuits and Systems for Video Technology, 36(7), 10721-10736. https://doi.org/10.1109/tcsvt.2026.3672454
Vancouver
Jiao Q, Shi M, Yang Y, Zhang Q. QANet: Query-Aware Multi-Modal Prior Refinement for Few-Shot Segmentation. IEEE Transactions on Circuits and Systems for Video Technology. 2026;36(7):10721-10736. doi:10.1109/tcsvt.2026.3672454
BibTeX
@article{Jiao2026,
author = {Jiao, Qiang and Shi, Mengrui and Yang, Yi and Zhang, Qiang},
title = {QANet: Query-Aware Multi-Modal Prior Refinement for Few-Shot Segmentation},
journal = {IEEE Transactions on Circuits and Systems for Video Technology},
year = {2026},
volume = {36},
number = {7},
pages = {10721--10736},
doi = {10.1109/tcsvt.2026.3672454}
}
EUVIMED