Modeling Pixel Means and Covariances Using Factorized Third-Order Boltzmann Machines

被引:82
|
作者
Ranzato, Marc'Aurelio [1 ]
Hinton, Geoffrey E. [1 ]
机构
[1] Univ Toronto, Dept Comp Sci, Toronto, ON, Canada
来源
2010 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR) | 2010年
关键词
IMAGES; SCENE; SET;
D O I
10.1109/CVPR.2010.5539962
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Learning a generative model of natural images is a useful way of extracting features that capture interesting regularities. Previous work on learning such models has focused on methods in which the latent features are used to determine the mean and variance of each pixel independently, or on methods in which the hidden units determine the covariance matrix of a zero-mean Gaussian distribution. In this work, we propose a probabilistic model that combines these two approaches into a single framework. We represent each image using one set of binary latent features that model the image-specific covariance and a separate set that model the mean. We show that this approach provides a probabilistic framework for the widely used simple-cell complex-cell architecture, it produces very realistic samples of natural images and it extracts features that yield state-of-the-art recognition accuracy on the challenging CIFAR 10 dataset.
引用
收藏
页码:2551 / 2558
页数:8
相关论文
共 50 条
  • [21] Adaptive detection of noisy speech using third-order statistics
    Natl Technical Univ of Athens, Athens, Greece
    Int J Adapt Control Signal Process, 2-3 (113-136):
  • [22] Third-order Volterra kernels complexity reduction using PARAFAC
    Khouaja, A
    Kibangou, A
    Favier, G
    ISCCSP : 2004 FIRST INTERNATIONAL SYMPOSIUM ON CONTROL, COMMUNICATIONS AND SIGNAL PROCESSING, 2004, : 857 - 860
  • [23] Blind equalization of PPM signals using third-order moments
    Paajarvi, Patrik
    LeBlanc, James P.
    2007 IEEE 8TH WORKSHOP ON SIGNAL PROCESSING ADVANCES IN WIRELESS COMMUNICATIONS, VOLS 1 AND 2, 2007, : 485 - 489
  • [24] Electronically Tunable Third-Order Quadrature Oscillator Using CDTAs
    Horng, Jiun-Wei
    Lee, Hung
    Wu, Jian-Yu
    RADIOENGINEERING, 2010, 19 (02) : 326 - 330
  • [25] Electronically Tunable Third-Order Quadrature Oscillator Using VDTAs
    Chen, Hua-Pin
    Wang, San-Fu
    Chen, Yu-Nan
    Huang, Qi-Geng
    JOURNAL OF CIRCUITS SYSTEMS AND COMPUTERS, 2019, 28 (04)
  • [26] Third-order dispersion compensation using 4 phase modulator
    Hellström, E
    Sunnerud, H
    Westlund, M
    Karlsson, M
    JOURNAL OF LIGHTWAVE TECHNOLOGY, 2003, 21 (05) : 1188 - 1197
  • [27] New Third-Order Oscillators Using Two or Three Amplifiers
    Elwakil, A. S.
    Maundy, B. J.
    Psychalinos, C.
    CIRCUITS SYSTEMS AND SIGNAL PROCESSING, 2024, 43 (08) : 4750 - 4763
  • [28] Third-order quadrature oscillator with grounded capacitors using CCIIs
    Horng, Jiun-Wei
    Hou, Chun-Li
    Chang, Chun-Ming
    Pan, Sheng-Vven
    Shie, Jhenyu
    Wen, Yao-Hsin
    6TH WSEAS INT CONF ON INSTRUMENTATION, MEASUREMENT, CIRCUITS & SYSTEMS/7TH WSEAS INT CONF ON ROBOTICS, CONTROL AND MANUFACTURING TECHNOLOGY, PROCEEDINGS, 2007, : 89 - +
  • [29] A method for MISO systems identification using third-order cumulants
    Korze, D
    Zazula, D
    Potocnik, B
    DSP 97: 1997 13TH INTERNATIONAL CONFERENCE ON DIGITAL SIGNAL PROCESSING PROCEEDINGS, VOLS 1 AND 2: SPECIAL SESSIONS, 1997, : 927 - 930
  • [30] Adaptive detection of noisy speech using third-order statistics
    Rangoussi, M
    Carayannis, G
    INTERNATIONAL JOURNAL OF ADAPTIVE CONTROL AND SIGNAL PROCESSING, 1996, 10 (2-3) : 113 - 136