Deep Learning-based Image Compression with Trellis Coded Quantization

TL;DR

This paper introduces a deep learning image compression model that integrates trellis coded quantization (TCQ), demonstrating improved performance over traditional scalar quantization at low bit rates through end-to-end training.

Contribution

It is the first to incorporate TCQ into a deep learning framework for image compression, enhancing compression efficiency at low bit rates.

Findings

TCQ outperforms scalar quantization in experiments

The model achieves superior performance on high-resolution datasets

End-to-end training effectively optimizes all components

Abstract

Recently many works attempt to develop image compression models based on deep learning architectures, where the uniform scalar quantizer (SQ) is commonly applied to the feature maps between the encoder and decoder. In this paper, we propose to incorporate trellis coded quantizer (TCQ) into a deep learning based image compression framework. A soft-to-hard strategy is applied to allow for back propagation during training. We develop a simple image compression model that consists of three subnetworks (encoder, decoder and entropy estimation), and optimize all of the components in an end-to-end manner. We experiment on two high resolution image datasets and both show that our model can achieve superior performance at low bit rates. We also show the comparisons between TCQ and SQ based on our proposed baseline model and demonstrate the advantage of TCQ.