Perceptually Motivated Method for Image Inpainting Comparison

TL;DR

This paper conducts a subjective comparison of nine image inpainting algorithms and proposes new objective quality metrics that better align with human perception of realism.

Contribution

It introduces a standardized evaluation method for image inpainting algorithms based on human perception and proposes metrics that correlate well with subjective assessments.

Findings

Proposed objective metrics align closely with human judgments

Subjective comparison reveals differences in algorithm realism

Standardized evaluation framework for future research

Abstract

The field of automatic image inpainting has progressed rapidly in recent years, but no one has yet proposed a standard method of evaluating algorithms. This absence is due to the problem's challenging nature: image-inpainting algorithms strive for realism in the resulting images, but realism is a subjective concept intrinsic to human perception. Existing objective image-quality metrics provide a poor approximation of what humans consider more or less realistic. To improve the situation and to better organize both prior and future research in this field, we conducted a subjective comparison of nine state-of-the-art inpainting algorithms and propose objective quality metrics that exhibit high correlation with the results of our comparison.