Reduced-Complexity End-to-End Variational Autoencoder for on Board Satellite Image Compression

Abstract

Recently, convolutional neural networks have been successfully applied to lossy image compression. End-to-end optimized autoencoders, possibly variational, are able to dramatically outperform traditional transform coding schemes in terms of rate-distortion trade-off; however, this is at the cost of a higher computational complexity. An intensive training step on huge databases allows autoencoders to learn jointly the image representation and its probability distribution, possibly using a non-parametric density model or a hyperprior auxiliary autoencoder to eliminate the need for prior knowledge. However, in the context of on board satellite compression, time and memory complexities are submitted to strong constraints. The aim of this paper is to design a complexity-reduced variational autoencoder in order to meet these constraints while maintaining the performance. Apart from a network dimension reduction that systematically targets each parameter of the analysis and synthesis transforms, we propose a simplified entropy model that preserves the adaptability to the input image. Indeed, a statistical analysis performed on satellite images shows that the Laplacian distribution fits most features of their representation. A complex non parametric distribution fitting or a cumbersome hyperprior auxiliary autoencoder can thus be replaced by a simple parametric estimation. The proposed complexity-reduced autoencoder outperforms the Consultative Committee for Space Data Systems standard (CCSDS 122.0-B) while maintaining a competitive performance, in terms of rate-distortion trade-off, in comparison with the state-of-the-art learned image compression schemes.

Deep Generative Learning-Based 1-SVM Detectors for Unsupervised COVID-19 Infection Detection Using Blood Tests

“…VAE, as a stochastic generative model, has become a popular modeling technique for learning underlying distributions of the input data by reconstruction and producing new data points based on the estimated distribution. Since its first introduction in 2014, the VAE method has been found helpful in numerous applications, such as time-series forecasting [30]- [32], anomaly detection [33]- [35], and image analysis [36].…”

Editorial to Special Issue “Remote Sensing Data Compression”

“…The paper "Reduced-Complexity End-to-End Variational Autoencoder for on Board Satellite Image Compression" by de Oliveira, V.A., Chabert, M., Oberlin, T., Poulliat, C., Bruno, M., Latry, C., Carlavan, M., Henrot, S., Falzon, F., and Camarero, R. [11] concentrates on design of a complexity-reduced variational autoencoder with attempt to meet the constraints dealing with board satellite compression, time, and memory complexities. A simplified entropy model that preserves the adaptability to the input image is proposed.…”

Satellite Image Compression and Denoising With Neural Networks

“…Recently, end-toend CNNs [17], [18] were shown to outperform traditional compression schemes regarding the rate-distortion trade-off, however, at the cost of high computational complexity. Based on the model proposed in [18], we presented in our previous paper [1] a satellite image compression variant with reduced complexity and competitive performance. Furthermore, denoising CNNs can adapt to any non-standard noise statistical model as soon as it can be learned from a representative training data set.…”

“…The second proposed approach sequentially uses a first neural architecture for onboard compression and a second one for on ground denoising. For both approaches, the onboard architectures are lightened as much as possible, following the procedure proposed in [1]. The two approaches are shown to outperform the current satellite imaging system and their respective pros and cons are discussed.…”