EUVIMEDEUVIMED entdecken
Crossref

HiRQA: Hierarchical Ranking and Quality Alignment for Opinion-Unaware Image Quality Assessment

Vaishnav Ramesh, Haining Wang, Md Jahidul Islam

Machine Vision and Applications · 2026-09 · 37 (5)

DOI: 10.1007/s00138-026-01845-6 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Machine Vision and Applications, online erschienen am 15.07.2026. Verfasst von Vaishnav Ramesh und 2 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 92 Literaturangaben. Gefördert von: National Science Foundation, University of Florida.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Despite significant progress in no-reference image quality assessment (NR-IQA), dataset biases and reliance on subjective labels continue to hinder their generalization performance. We propose HiRQA (Hierarchical Ranking and Quality Alignment), a self-supervised, opinion-unaware framework that offers a hierarchical, quality-aware embedding through a combination of ranking and contrastive learning. Unlike prior approaches that depend on pristine references or auxiliary modalities at inference time, HiRQA predicts quality scores using only the input image. We introduce a novel higher-order ranking loss that supervises quality predictions through relational ordering across distortion pairs, along with an embedding distance loss that enforces consistency between feature distances and perceptual differences. A training-time contrastive alignment loss, guided by structured textual prompts, further enhances the learned representation. Trained only on synthetic image distortions, HiRQA generalizes to authentic degradations, as demonstrated through comprehensive evaluations on various unseen distortions such as lens flare, haze, motion blur, and low-light conditions. For real-time deployment, we introduce HiRQA-S, a lightweight variant with an inference time of only 3.5 ms per image. Extensive experiments across synthetic and authentic benchmarks validate HiRQA’s competitive performance, strong generalization ability, and scalability. The HiRQA model and inference pipeline are available at: https://github.com/uf-robopi/HiRQA.

Förderung

  • National Science Foundation
  • University of Florida

Zitieren

APA

Ramesh, V., Wang, H., & Islam, M. J. (2026). HiRQA: Hierarchical Ranking and Quality Alignment for Opinion-Unaware Image Quality Assessment. Machine Vision and Applications, 37(5). https://doi.org/10.1007/s00138-026-01845-6

Vancouver

Ramesh V, Wang H, Islam MJ. HiRQA: Hierarchical Ranking and Quality Alignment for Opinion-Unaware Image Quality Assessment. Machine Vision and Applications. 2026;37(5). doi:10.1007/s00138-026-01845-6

BibTeX
@article{Ramesh2026,
  author = {Ramesh, Vaishnav and Wang, Haining and Islam, Md Jahidul},
  title = {HiRQA: Hierarchical Ranking and Quality Alignment for Opinion-Unaware Image Quality Assessment},
  journal = {Machine Vision and Applications},
  year = {2026},
  volume = {37},
  number = {5},
  doi = {10.1007/s00138-026-01845-6}
}
Kostenlos starten →