Image Modeling with Deep Convolutional Gaussian Mixture Models

被引:0
作者
Gepperth, Alexander [1 ]
Pfuelb, Benedikt [1 ]
机构
[1] Fulda Univ Appl Sci, Fulda, Germany
来源
2021 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN) | 2021年
关键词
Deep Learning; Gaussian Mixture Models; Deep Convolutional Gaussian Mixture Models; Stochastic Gradient Descent;
D O I
10.1109/IJCNN52387.2021.9533745
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In this conceptual work, we present Deep Convolutional Gaussian Mixture Models (DCGMMs): a new formulation of deep hierarchical Gaussian Mixture Models (GMMs) that is particularly suitable for describing and generating images. Vanilla (i.e., flat) GMMs require a very large number of components to describe images well, leading to long training times and memory issues. DCGMMs avoid this by a stacked architecture of multiple GMM layers, linked by convolution and pooling operations. This allows to exploit the compositionality of images in a similar way as deep CNNs do. DCGMMs can be trained end-to-end by Stochastic Gradient Descent. This sets them apart from vanilla GMMs which are trained by Expectation-Maximization, requiring a prior k-means initialization which is infeasible in a layered structure. For generating sharp images with DCGMMs, we introduce a new gradient-based technique for sampling through non-invertible operations like convolution and pooling. Based on the MNIST and FashionMNIST datasets, we validate the DCGMMs model by demonstrating its superiority over flat GMMs for clustering, sampling and outlier detection.
引用
收藏
页数:9
相关论文
共 50 条
  • [31] A Novel Gaussian-Bernoulli Based Convolutional Deep Belief Networks for Image Feature Extraction
    Li, Ziqiang
    Cai, Xun
    Liu, Yun
    Zhu, Bo
    NEURAL PROCESSING LETTERS, 2019, 49 (01) : 305 - 319
  • [32] Baseline building energy modeling and localized uncertainty quantification using Gaussian mixture models
    Srivastav, Abhishek
    Tewari, Ashutosh
    Dong, Bing
    ENERGY AND BUILDINGS, 2013, 65 : 438 - 447
  • [33] Contact-State Modeling of Robotic Assembly Tasks Using Gaussian Mixture Models
    Jasim, Ibrahim F.
    Plapper, Peter W.
    5TH CATS 2014 - CIRP CONFERENCE ON ASSEMBLY TECHNOLOGIES AND SYSTEMS, 2014, 23 : 229 - 234
  • [34] Direct Importance Estimation with Gaussian Mixture Models
    Yamada, Makoto
    Sugiyama, Masashi
    IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS, 2009, E92D (10) : 2159 - 2162
  • [35] Fitting Gaussian mixture models on incomplete data
    Zachary R. McCaw
    Hugues Aschard
    Hanna Julienne
    BMC Bioinformatics, 23
  • [36] SUBSPACE GAUSSIAN MIXTURE MODELS FOR SPEECH RECOGNITION
    Povey, Daniel
    Burget, Lukas
    Agarwal, Mohit
    Akyazi, Pinar
    Feng, Kai
    Ghoshal, Arnab
    Glembek, Ondrej
    Goel, Nagendra Kumar
    Karafiat, Martin
    Rastrow, Ariya
    Rose, Richard C.
    Schwarz, Petr
    Thomas, Samuel
    2010 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, 2010, : 4330 - 4333
  • [37] Fitting Gaussian mixture models on incomplete data
    McCaw, Zachary R.
    Aschard, Hugues
    Julienne, Hanna
    BMC BIOINFORMATICS, 2022, 23 (01)
  • [38] Statistical Compressed Sensing of Gaussian Mixture Models
    Yu, Guoshen
    Sapiro, Guillermo
    IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2011, 59 (12) : 5842 - 5858
  • [39] Cutoff for Exact Recovery of Gaussian Mixture Models
    Chen, Xiaohui
    Yang, Yun
    IEEE TRANSACTIONS ON INFORMATION THEORY, 2021, 67 (06) : 4223 - 4238
  • [40] Local Minima Structures in Gaussian Mixture Models
    Chen, Yudong
    Song, Dogyoon
    Xi, Xumei
    Zhang, Yuqian
    IEEE TRANSACTIONS ON INFORMATION THEORY, 2024, 70 (06) : 4218 - 4257