Perceptual no-reference image quality assessment with meta-learning by graph representation learning and multi-scale feature fusion

Yanru Jia, Longsheng Wei · 2026

Read Perceptual no-reference image quality assessment with meta-learning by graph representation learning and multi-scale feature fusion by Yanru Jia, Longsheng Wei (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0351549.txt.

Description

Rating how good a photo looks, without a clean original to compare it with, is the problem Yanru Jia and Longsheng Wei take on in this 2026 PLOS One paper. Phone cameras and sharing apps need such automatic scores, but blur, noise, and compression each damage pictures in different ways.

Their method first trains a vision network that adjusts itself, then arranges sample images in a graph linking damage type, damage level, and picture content. On four public test collections the scores lined up with human judgments better than older hand-built and deep learning methods. Tests removing one part at a time showed each piece added accuracy.

Catalog record

CREATORYanru Jia, Longsheng Wei
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE49770
LICENSECC_BY
RIGHTSCC BY
SOURCEPLOS

Source: https://doi.org/10.1371/journal.pone.0351549

Browse the full library · About the Tower of Babel