Abstract:Recent deep image compression methods have achieved prominent progress by using nonlinear modeling and powerful representation capabilities of neural networks. However, most existing learning-based image compression approaches employ customized convolutional neural network (CNN) to utilize visual features by treating all pixels equally, neglecting the effect of local key features. Meanwhile, the convolutional filters in CNN usually express the local spatial relationship within the receptive field and seldom consider the long-range dependencies from distant locations. This results in the long-range dependencies of latent representations not being fully compressed. To address these issues, an end-to-end image compression method is proposed by integrating graph attention and asymmetric convolutional neural network (ACNN). Specifically, ACNN is used to strengthen the effect of local key features and reduce the cost of model training. Graph attention is introduced into image compression to address the bottleneck problem of CNN in modeling long-range dependencies. Meanwhile, regarding the limitation that existing attention mechanisms for image compression hardly share information, we propose a self-attention approach which allows information flow to achieve reasonable bit allocation. The proposed self-attention approach is in compliance with the perceptual characteristics of human visual system, as information can interact with each other via attention modules. Moreover, the proposed self-attention approach takes into account channel-level relationship and positional information to promote the compression effect of rich-texture regions. Experimental results demonstrate that the proposed method achieves state-of-the-art rate-distortion performances after being optimized by MS-SSIM compared to recent deep compression models on the benchmark datasets of Kodak and Tecnick. The project page with the source code can be found in https://mic.tongji.edu.cn.

Dynamic Neural Networks for Adaptive Implicit Image Compression.

COIN++: Neural Compression Across Modalities

Implicit Neural Image Field for Biological Microscopy Image Compression

Content Adaptive Latents and Decoder for Neural Image Compression

Breaking the Barriers of One-to-One Usage of Implicit Neural Representation in Image Compression: A Linear Combination Approach with Performance Guarantees

Extreme Compression of Adaptive Neural Images

Dynamic Low-Rank Instance Adaptation for Universal Neural Image Compression

Approximately Invertible Neural Network for Learned Image Compression

Computationally-Efficient Neural Image Compression with Shallow Decoders

Hybrid Implicit Neural Image Compression with Subpixel Context Model and Iterative Pruner

Learned Image Compression for Both Humans and Machines Via Dynamic Adaptation

Mixed-Resolution Image Representation and Compression with Convolutional Neural Networks.

A Unified End-to-End Framework for Efficient Deep Image Compression

Joint Graph Attention and Asymmetric Convolutional Neural Network for Deep Image Compression

Facial Image Compression via Neural Image Manifold Compression

Towards On-demand Transmission: Joint Feature and Image Coding with Reversible Neural Networks

Spatially adaptive image compression using a tiled deep network

A Novel Deep Progressive Image Compression Framework

Compression with Bayesian Implicit Neural Representations

Implicit Neural Video Compression

Content Adaptive Optimization for Neural Image Compression