Lossy Image Compression with Compressive Autoencoders

Abstract

We propose a new approach to the problem of optimizing autoencoders for lossy image compression. New media formats, changing hardware technology, as well as diverse requirements and content types create a need for compression algorithms which are more flexible than existing codecs. Autoencoders have the potential to address this need, but are difficult to optimize directly due to the inherent non-differentiabilty of the compression loss. We here show that minimal changes to the loss are sufficient to train deep autoencoders competitive with JPEG 2000 and outperforming recently proposed approaches based on RNNs. Our network is furthermore computationally efficient thanks to a sub-pixel architecture, which makes it suitable for high-resolution images. This is in contrast to previous work on autoencoders for compression using coarser approximations, shallower architectures, computationally expensive methods, or focusing on small images.

References in corpus (2)

• scikit-image: Image processing in Python
• Auto-encoders: reconstruction versus compression

Cited by in corpus (15)

• Image and Video Compression with Neural Networks: A Review
• Real-Time Adaptive Image Compression
• Learning to Inpaint for Image Compression
• Improved Lossy Image Compression with Priming and Spatially Adaptive Bit Rates for Recurrent Networks
• Regularizing linear inverse problems with convolutional neural networks
• LatentPoison - Adversarial Attacks On The Latent Space
• GAN- vs. JPEG2000 Image Compression for Distributed Automotive Perception: Higher Peak SNR Does Not Mean Better Semantic Segmentation
• Gated Context Model with Embedded Priors for Deep Image Compression
• CAE-ADMM: Implicit Bitrate Optimization via ADMM-based Pruning in Compressive Autoencoders
• Toward Joint Image Generation and Compression using Generative Adversarial Networks
• Attention Based Image Compression Post-Processing Convolutional Neural Network
• Learning to compress and search visual data in large-scale systems
• Layered Image Compression using Scalable Auto-encoder
• A Deep Image Compression Framework for Face Recognition
• DeepSIC: Deep Semantic Image Compression