Advancements in Image Quality Assessment and Restoration

The field of image quality assessment and restoration is moving towards a more nuanced understanding of human visual perception, with a focus on multi-dimensional evaluation and restoration. Researchers are developing frameworks that capture the multifaceted nature of human visual perception, including technical and aesthetic dimensions. These frameworks are being applied to various tasks, such as image restoration, denoising, and license plate recognition. Notable papers in this area include: MDIQA, which proposes a unified image quality assessment framework for multi-dimensional evaluation and restoration, achieving superior performance and flexibility in image restoration tasks. VQualA 2025 Challenge, which summarizes methodologies and findings for advancing the development of practical face image quality assessment approaches, attracting 127 participants and 1519 final submissions.

Sources

MDIQA: Unified Image Quality Assessment for Multi-dimensional Evaluation and Restoration

AIM 2025 Low-light RAW Video Denoising Challenge: Dataset, Methods and Results

LPLC: A Dataset for License Plate Legibility Classification

VQualA 2025 Challenge on Face Image Quality Assessment: Methods and Results

Image Quality Assessment for Machines: Paradigm, Large-scale Database, and Models

Built with on top of