EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Visual–Label Alignment and Attribute-Aware Prompt for Multi-Label Image Recognition with Partial Labels

Dengdi Sun, Hongxing Xie, Zhendong Cai, Leilei Ma, Bin Luo

ACM Transactions on Multimedia Computing, Communications, and Applications · 31.08.2026 · 22 (8) · 1-24

DOI: 10.1145/3828542 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Transactions on Multimedia Computing, Communications, and Applications, online erschienen am 21.07.2026. Verfasst von Dengdi Sun und 4 weiteren Autor:innen. Die Arbeit nennt 55 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: National Natural Science Foundation of China, Dreams Foundation of Jianghuai Advance Technology Center, Natural Science Foundation of Anhui Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The problem of Multi-Label Image Recognition with Partial Labels (MLIR-PL) is a significant challenge in computer vision, primarily due to the scarcity and high cost of complete annotations. Recent advances have leveraged large-scale vision-language models, such as CLIP, to establish rich correspondences between images and their labels, thereby improving the MLIR-PL performance. However, the existing CLIP-based methods have not fully exploited fine-grained local image features to mitigate interference from semantically irrelevant regions. Moreover, many studies have oversimplified the use of prompt contexts, limiting their ability to comprehensively capture the multi-dimensional attributes of categories. To address these limitations, this article proposes a novel MLIR-PL model with Visual–Label Alignment and Attribute-Aware Prompt (VA \({}^{3}\) P), which sufficiently harnesses the capabilities of large-scale pre-trained vision-language models. In the model, we design a Visual–Label Alignment module to establish a mapping between local image features and category text representations, conspicuously reducing the interference from irrelevant regions. Additionally, our Attribute-Aware Prompt module offers diverse contextual information, providing a more comprehensive representation of the category’s attributes. Extensive experimental results on the COCO 2014 and VOC 2007 datasets, compared with multiple state-of-the-art methods, demonstrate that our model achieves the best performance comprehensively, verifying the advantages of the proposed model in the MLIR-PL task.

Förderung

  • National Natural Science Foundation of China
  • Dreams Foundation of Jianghuai Advance Technology Center
  • Natural Science Foundation of Anhui Province
  • Anhui Provincial University Natural Science Research Major Project
  • University Synergy Innovation Program of Anhui Province

Zitieren

APA

Sun, D., Xie, H., Cai, Z., Ma, L., & Luo, B. (2026). Visual–Label Alignment and Attribute-Aware Prompt for Multi-Label Image Recognition with Partial Labels. ACM Transactions on Multimedia Computing, Communications, and Applications, 22(8), 1-24. https://doi.org/10.1145/3828542

Vancouver

Sun D, Xie H, Cai Z, Ma L, Luo B. Visual–Label Alignment and Attribute-Aware Prompt for Multi-Label Image Recognition with Partial Labels. ACM Transactions on Multimedia Computing, Communications, and Applications. 2026;22(8):1-24. doi:10.1145/3828542

BibTeX
@article{Sun2026,
  author = {Sun, Dengdi and Xie, Hongxing and Cai, Zhendong and Ma, Leilei and Luo, Bin},
  title = {Visual–Label Alignment and Attribute-Aware Prompt for Multi-Label Image Recognition with Partial Labels},
  journal = {ACM Transactions on Multimedia Computing, Communications, and Applications},
  year = {2026},
  volume = {22},
  number = {8},
  pages = {1--24},
  doi = {10.1145/3828542}
}
Kostenlos starten →