Advances in Image Quality Assessment and Document Dewarping

The field of image processing is moving towards more sophisticated and nuanced methods for evaluating and improving image quality. This includes the development of new metrics and frameworks for assessing image quality, such as those that incorporate region-aware semantic attention and anisotropic texture richness. Additionally, there is a growing focus on addressing the challenges of document image dewarping, including the development of time-aware models and novel evaluation metrics. These advances have the potential to significantly improve the accuracy and robustness of image quality assessment and document dewarping systems. Noteworthy papers include: TADoc, which proposes a time-aware document dewarping network and a new metric for evaluating document layout similarity. The Parametric Bi-Directional Curvature-Based Framework, which presents a novel framework for image artifact classification and quantification based on the analysis of directional image curvature.

Sources

TADoc: Robust Time-Aware Document Image Dewarping

Segmenting and Understanding: Region-aware Semantic Attention for Fine-grained Image Quality Assessment with Large Language Models

A Parametric Bi-Directional Curvature-Based Framework for Image Artifact Classification and Quantification

In-place Double Stimulus Methodology for Subjective Assessment of High Quality Images

Hierarchical Graph Attention Network for No-Reference Omnidirectional Image Quality Assessment

Built with on top of