MM-IQA combines hand-crafted cues for common image distortions into a single lightweight NR-IQA score, reporting SRCC values of 0.647-0.830 on five standard benchmarks and running in under 2 seconds per image.
Koniq-10k: An ecologically valid database for deep learning of blind image quality assessment.IEEE Access, 8:121257–121270
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Lightweight Multi-Metric No-Reference Image Quality Assessment Framework for UAV Imaging
MM-IQA combines hand-crafted cues for common image distortions into a single lightweight NR-IQA score, reporting SRCC values of 0.647-0.830 on five standard benchmarks and running in under 2 seconds per image.