Convolutional Fully-Connected Capsule Network (CFC-CapsNet): A Novel and Fast Capsule Network

被引:6
|
作者
Shiri, Pouya [1 ]
Baniasadi, Amirali [1 ]
机构
[1] Univ Victoria, Elect & Comp Engn Fac, Victoria, BC, Canada
来源
JOURNAL OF SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY | 2022年 / 94卷 / 07期
基金
加拿大自然科学与工程研究理事会;
关键词
Capsule Networks; CapsNet; Deep Learning; Fast CapsNet;
D O I
10.1007/s11265-021-01731-6
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
A Capsule Network (CapsNet) is a relatively new classifier and one of the possible successors of Convolutional Neural Networks (CNNs). CapsNet maintains the spatial hierarchies between the features and outperforms CNNs at classifying images including overlapping categories. Even though CapsNet works well on small-scale datasets such as MNIST, it fails to achieve a similar level of performance on more complicated datasets and real applications. In addition, CapsNet is slow compared to CNNs when performing the same task and relies on a higher number of parameters. In this work, we introduce Convolutional Fully-Connected Capsule Network (CFC-CapsNet) to address the shortcomings of CapsNet by creating capsules using a different method. We introduce a new layer (CFC layer) as an alternative solution to creating capsules. CFC-CapsNet produces fewer, yet more powerful capsules resulting in higher network accuracy. Our experiments show that CFC-CapsNet achieves competitive accuracy, faster training and inference and uses less number of parameters on the CIFAR-10, SVHN and Fashion-MNIST datasets compared to conventional CapsNet.
引用
收藏
页码:645 / 658
页数:14
相关论文
共 50 条
  • [1] Convolutional Fully-Connected Capsule Network (CFC-CapsNet): A Novel and Fast Capsule Network
    Pouya Shiri
    Amirali Baniasadi
    Journal of Signal Processing Systems, 2022, 94 : 645 - 658
  • [2] Skin Cancer Classification using Convolutional Capsule Network (CapsNet)
    Cruz, Meenalosini Vimal
    Namburu, Anupama
    Chakkaravarthy, Sibi
    Pittendreigh, Matthew
    Satapathy, Suresh Chandra
    JOURNAL OF SCIENTIFIC & INDUSTRIAL RESEARCH, 2020, 79 (11): : 994 - 1001
  • [3] RS-CapsNet: An Advanced Capsule Network
    Yang, Shuai
    Lee, Feifei
    Miao, Ran
    Cai, Jiawei
    Chen, Lu
    Yao, Wei
    Kotani, Koji
    Chen, Qiu
    IEEE ACCESS, 2020, 8 : 85007 - 85018
  • [4] MS-CapsNet: A Novel Multi-Scale Capsule Network
    Xiang, Canqun
    Zhang, Lu
    Tang, Yi
    Zou, Wenbin
    Xu, Chen
    IEEE SIGNAL PROCESSING LETTERS, 2018, 25 (12) : 1850 - 1854
  • [5] FSC-CapsNet: Fractionally-Strided Convolutional Capsule Network for complex data
    Liu, Jian-wei
    Gao, Feng
    Lu, Run-kun
    Lian, Yuan-feng
    Wang, Dian-zhong
    Luo, Xiong-lin
    Wang, Chu-ran
    2019 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2019,
  • [6] DL-CapsNet: A Deep and Light Capsule Network
    Shiri, Pouya
    Baniasadi, Amirali
    DESIGN AND ARCHITECTURE FOR SIGNAL AND IMAGE PROCESSING, DASIP 2022, 2022, 13425 : 57 - 68
  • [7] LE-CapsNet: A Light and Enhanced Capsule Network
    Shiri, Pouya
    Baniasadi, Amirali
    20TH IEEE INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND APPLICATIONS (ICMLA 2021), 2021, : 1767 - 1772
  • [8] CAT-CapsNet: A Convolutional and Attention Based Capsule Network to Detect the Driver's Distraction
    Mittal, Himanshu
    Verma, Bindu
    IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, 2023, 24 (09) : 9561 - 9570
  • [9] H-CapsNet: A capsule network for hierarchical image classification
    Noor, Khondaker Tasrif
    Robles-Kelly, Antonio
    PATTERN RECOGNITION, 2024, 147
  • [10] Res-CapsNet: Residual Capsule Network for Data Classification
    Jia, Xiaofen
    Li, Jianqiao
    Zhao, Baiting
    Guo, Yongcun
    Huang, Yourui
    NEURAL PROCESSING LETTERS, 2022, 54 (05) : 4229 - 4245