Paper Info
Title
Noise-to-Compression Variational Autoencoder for Efficient End-to-End Optimized Image Coding
Abstract
Generative model has emerged as a disruptive alternative for lossy compression of natural images, but suffers from the low-fidelity reconstruction. In this paper, we propose a noise-to-compression variational antoencoder (NC-VAE) to achieve efficient rate-distortion optimization (RDO) for end-to-end optimized image compression with a guarantee of fidelity. The proposed NC-VAE improves rate-distortion performance by adaptively adjusting the distribution of latent variables with trainable noise perturbation. Consequently, high-efficiency RDO is developed based on the distribution of latent variables for simplified decoder. Furthermore, robust end-to-end learning is developed over the corrupted inputs to suppress the deformation and color drift in standard VAE based generative models. Experimental results show that NC-VAE outperforms the state-of-the-art lossy image coders and recent end-to-end optimized compression methods in low bit-rate region, i.e., below 0.2 bits per pixel (bpp).
Year
DOI
Venue
2020
10.1109/DCC47342.2020.00011
2020 Data Compression Conference (DCC)
Keywords
DocType
ISSN
color drift,efficient end-to-end optimized image coding,high-efficiency RDO,trainable noise perturbation,end-to-end optimized image compression,rate-distortion optimization,noise-to-compression variational antoencoder,low-fidelity reconstruction,natural images,lossy compression,noise-to-compression variational autoencoder,low bit-rate region,compression methods,state-of-the-art lossy image coders,NC-VAE,standard VAE based generative models,robust end-to-end learning
Conference
1068-0314
ISBN
Citations 
PageRank 
978-1-7281-6458-8
1
0.37
References 
Authors
4
7
Name
Order
Citations
PageRank
Jixiang Luo121.73
Shaohui Li2203.75
Wenrui Dai36425.01
Yuhui Xu4125.00
De Cheng5707.97
Gang Li632.82
Hongkai Xiong7228.85