Analysis of Neural Image Compression Networks for Machine-to-Machine Communication

TL;DR

This paper evaluates four state-of-the-art neural compression networks for machine-to-machine image communication, showing that certain architectures and training methods significantly improve coding performance, with some surpassing traditional standards like VVC.

Contribution

It provides a comprehensive evaluation framework for neural image compression networks in machine-to-machine scenarios, highlighting the impact of architecture choices and training criteria on performance.

Findings

Networks with leaky ReLU and SSIM training perform best.

GAN-based NCN architecture outperforms VVC in the scenario.

Neural compression networks can surpass traditional codecs for machine analysis tasks.

Abstract

Video and image coding for machines (VCM) is an emerging field that aims to develop compression methods resulting in optimal bitstreams when the decoded frames are analyzed by a neural network. Several approaches already exist improving classic hybrid codecs for this task. However, neural compression networks (NCNs) have made an enormous progress in coding images over the last years. Thus, it is reasonable to consider such NCNs, when the information sink at the decoder side is a neural network as well. Therefore, we build-up an evaluation framework analyzing the performance of four state-of-the-art NCNs, when a Mask R-CNN is segmenting objects from the decoded image. The compression performance is measured by the weighted average precision for the Cityscapes dataset. Based on that analysis, we find that networks with leaky ReLU as non-linearity and training with SSIM as distortion criteria results in the highest coding gains for the VCM task. Furthermore, it is shown that the GAN-based NCN architecture achieves the best coding performance and even out-performs the recently standardized Versatile Video Coding (VVC) for the given scenario.