End-to-end optimized image compression with competition of prior distributions

被引:3
|
作者
Brummer, Benoit [1 ]
De Vleeschouwer, Christophe [2 ]
机构
[1] IntoPIX, Mont St Guibert, Belgium
[2] Catholic Univ Louvain, Louvain, Belgium
关键词
D O I
10.1109/CVPRW53098.2021.00212
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Convolutional autoencoders are now at the forefront of image compression research. To improve their entropy coding, encoder output is typically analyzed with a second autoencoder to generate per-variable parametrized prior probability distributions. We instead propose a compression scheme that uses a single convolutional autoencoder and multiple learned prior distributions working as a competition of experts. Trained prior distributions are stored in a static table of cumulative distribution functions. During inference, this table is used by an entropy coder as a look-up-table to determine the best prior for each spatial location. Our method offers rate-distortion performance comparable to that obtained with a predicted parametrized prior with only a fraction of its entropy coding and decoding complexity.
引用
收藏
页码:1890 / 1894
页数:5
相关论文
共 50 条
  • [1] End-to-End Optimized ROI Image Compression
    Cai, Chunlei
    Chen, Li
    Zhang, Xiaoyun
    Gao, Zhiyong
    [J]. IEEE TRANSACTIONS ON IMAGE PROCESSING, 2020, 29 : 3442 - 3457
  • [2] End-to-End Optimized 360° Image Compression
    Li, Mu
    Li, Jinxing
    Gu, Shuhang
    Wu, Feng
    Zhang, David
    [J]. IEEE TRANSACTIONS ON IMAGE PROCESSING, 2022, 31 : 6267 - 6281
  • [3] End-to-end optimized image compression for machines, a study
    Chamain, Lahiru D.
    Racape, Fabien
    Begaint, Jean
    Pushparaja, Akshay
    Feltman, Simon
    [J]. 2021 DATA COMPRESSION CONFERENCE (DCC 2021), 2021, : 163 - 172
  • [4] End-to-end optimized image compression with the frequency-oriented transform
    Yuefeng Zhang
    Kai Lin
    [J]. Machine Vision and Applications, 2024, 35
  • [5] End-to-end optimized image compression with the frequency-oriented transform
    Zhang, Yuefeng
    Lin, Kai
    [J]. MACHINE VISION AND APPLICATIONS, 2024, 35 (02)
  • [6] End-to-End Optimized Versatile Image Compression With Wavelet-Like Transform
    Ma, Haichuan
    Liu, Dong
    Yan, Ning
    Li, Houqiang
    Wu, Feng
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2022, 44 (03) : 1247 - 1263
  • [7] Volumetric End-to-End Optimized Compression for Brain Images
    Gao, Shuo
    Zhang, Yueyi
    Liu, Dong
    Xiong, Zhiwei
    [J]. 2020 IEEE INTERNATIONAL CONFERENCE ON VISUAL COMMUNICATIONS AND IMAGE PROCESSING (VCIP), 2020, : 503 - 506
  • [8] Noise-to-Compression Variational Autoencoder for Efficient End-to-End Optimized Image Coding
    Luo, Jixiang
    Li, Shaohui
    Dai, Wenrui
    Xu, Yuhui
    Cheng, De
    Li, Gang
    Xiong, Hongkai
    [J]. 2020 DATA COMPRESSION CONFERENCE (DCC 2020), 2020, : 33 - 42
  • [9] Efficient end-to-end multispectral image compression
    Depoian, Arthur C., II
    Bailey, Colleen P.
    Guturu, Parthasarathy
    [J]. BIG DATA VI: LEARNING, ANALYTICS, AND APPLICATIONS, 2024, 13036
  • [10] End-to-End Deep ROI Image Compression
    Akutsu, Hiroaki
    Naruko, Takahiro
    [J]. IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS, 2020, E103D (05): : 1031 - 1038