Comparison of Full-Reference Image Quality Models for Optimization of Image Processing Systems

The performance of objective image quality assessment (IQA) models has been evaluated primarily by comparing model predictions to human quality judgments. Perceptual datasets gathered for this purpose have provided useful benchmarks for improving IQA methods, but their heavy use creates a risk of overfitting. Here, we perform a large-scale comparison of IQA models in terms of their use as objectives for the optimization of image processing algorithms. Specifically, we use eleven full-reference IQA models to train deep neural networks for four low-level vision tasks: denoising, deblurring, super-resolution, and compression. Subjective testing on the optimized images allows us to rank the competing models in terms of their perceptual performance, elucidate their relative advantages and disadvantages in these tasks, and propose a set of desirable properties for incorporation into future IQA models.

Automated summary

The study compared eleven IQA models in terms of their effectiveness as objectives for optimizing image processing algorithms. Subjective testing allowed the researchers to rank the models based on perceptual performance, identify their relative strengths and weaknesses in low-level vision tasks, and propose desirable properties for future IQA models.