A Machine Learning Approach to Optimal Inverse Discrete Cosine Transform (IDCT) Design

TL;DR

This paper introduces a machine learning-based method to design an optimal inverse DCT that better compensates for quantization errors in lossy image compression, improving image quality over standard JPEG.

Contribution

A novel IDCT kernel is learned using training images to reverse quantization effects, enhancing compression quality beyond traditional methods.

Findings

Achieves 0.11-0.30dB gain over standard JPEG

Effectively compensates for quantization errors

Improves image quality across various quality factors

Abstract

The design of the optimal inverse discrete cosine transform (IDCT) to compensate the quantization error is proposed for effective lossy image compression in this work. The forward and inverse DCTs are designed in pair in current image/video coding standards without taking the quantization effect into account. Yet, the distribution of quantized DCT coefficients deviate from that of original DCT coefficients. This is particularly obvious when the quality factor of JPEG compressed images is small. To address this problem, we first use a set of training images to learn the compound effect of forward DCT, quantization and dequantization in cascade. Then, a new IDCT kernel is learned to reverse the effect of such a pipeline. Experiments are conducted to demonstrate that the advantage of the new method, which has a gain of 0.11-0.30dB over the standard JPEG over a wide range of quality factors.