Perceptually Optimized Loss Function for Image Super-Resolution

Most of the learning based single image super-resolution networks employ intensity loss which measures pixel-wise difference between the estimated high resolution image and the ground truth. Since image components are different with respect to their saliency for HVS, it is desired to weight their im...

Full description

Saved in:
Bibliographic Details
Published in2021 7th International Conference on Signal Processing and Intelligent Systems (ICSPIS) pp. 01 - 05
Main Authors Arezoomand, Amirhossein, Cheraaqee, Poorya, Mansouri, Azadeh
Format Conference Proceeding
LanguageEnglish
Published IEEE 29.12.2021
Subjects
Computer architecture
DCT
deep learning
Discrete cosine transforms
Energy resolution
JPEG
loss function
Loss measurement
neural networks
Quantization (signal)
single image super-resolution
Superresolution
Transform coding
Online AccessGet full text
DOI10.1109/ICSPIS54653.2021.9729334

Cover

More Information
Summary:Most of the learning based single image super-resolution networks employ intensity loss which measures pixel-wise difference between the estimated high resolution image and the ground truth. Since image components are different with respect to their saliency for HVS, it is desired to weight their impact on the loss functions accordingly. In this paper, a simple perceptual loss function is introduced based on the JPEG compression algorithm. In fact, the two compared images are transformed into DCT domain and then divided by the weighted quantization matrix. The difference between the resultant DCT coefficients shows the most effective components for HVS and can be considered as a perceptual loss function. The experimental results illustrate that employing the proposed loss promotes the convergence speed, and also, provides better outputs in terms of qualitative and quantitative measures.
DOI:10.1109/ICSPIS54653.2021.9729334