Geometric Back-Propagation in Morphological Neural Networks

被引：3

作者：

Groenendijk, Rick ^{[1
]}

Dorst, Leo ^{[1
]}

Gevers, Theo ^{[1
]}

机构：

[1] Univ Amsterdam, Comp Vis Grp, NL-1012 WP Amsterdam, Netherlands

来源：

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE | 2023年 / 45卷 / 11期

基金：

荷兰研究理事会;

关键词：

Neural networks; Probes; Convolution; Geometry; Task analysis; Morphological operations; Training; Mathematical morphology; morphological neural networks; back-propagation; probe geometry; depth infilling; RECONSTRUCTION; RESTORATION;

D O I：

10.1109/TPAMI.2023.3290615

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

This paper provides a definition of back-propagation through geometric correspondences for morphological neural networks. In addition, dilation layers are shown to learn probe geometry by erosion of layer inputs and outputs. A proof-of-principle is provided, in which predictions and convergence of morphological networks significantly outperform convolutional networks.

引用

页码：14045 / 14051

页数：7

共 50 条

[1] DESIGN OPTIMIZATION WITH BACK-PROPAGATION NEURAL NETWORKS
LEE, SJ
HENZER, C
[J]. JOURNAL OF INTELLIGENT MANUFACTURING, 1991, 2 (05) : 293 - 303
[2] GENERALIZATION OF BACK-PROPAGATION TO RECURRENT NEURAL NETWORKS
PINEDA, FJ
[J]. PHYSICAL REVIEW LETTERS, 1987, 59 (19) : 2229 - 2232
[3] Irregular shapes classification by back-propagation neural networks
Shih-Wei Lin
Shuo-Yan Chou
Shih-Chieh Chen
[J]. The International Journal of Advanced Manufacturing Technology, 2007, 34 : 1164 - 1172
[4] Irregular shapes classification by back-propagation neural networks
Lin, Shih-Wei
Chou, Shuo-Yan
Chen, Shih-Chieh
[J]. INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY, 2007, 34 (11-12): : 1164 - 1172
[5] Applying back-propagation neural networks to GDOP approximation
Jwo, DJ
Chin, KP
[J]. JOURNAL OF NAVIGATION, 2002, 55 (01): : 97 - 108
[6] Back-propagation Now Works in Spiking Neural Networks!
Masquelier, Timothee
[J]. ERCIM NEWS, 2021, (125): : 11 - 12
[7] THE UNPREDICTABILITY OF STANDARD BACK-PROPAGATION NEURAL NETWORKS IN CLASSIFICATION APPLICATIONS
WANG, SH
[J]. MANAGEMENT SCIENCE, 1995, 41 (03) : 555 - 559
[8] Using Back-propagation Neural Networks for Functional Software Testing
Wu, Lilan
Liu, Bo
Jin, Yi
Xie, Xiaoyao
[J]. 2008 2ND INTERNATIONAL CONFERENCE ON ANTI-COUNTERFEITING, SECURITY AND IDENTIFICATION, 2008, : 272 - +
[9] Generalized back-propagation algorithm with weight evolution for neural networks
Ng, SC
Leung, SH
Luk, A
[J]. 6TH WORLD MULTICONFERENCE ON SYSTEMICS, CYBERNETICS AND INFORMATICS, VOL XX, PROCEEDINGS EXTENSION, 2002, : 41 - 44
[10] Typhoon surge forecasting with artificial back-propagation neural networks
Jan, C. D.
Tseng, C. M.
Wang, J. S.
Wang, C. M.
[J]. OCEANS 2006 - ASIA PACIFIC, VOLS 1 AND 2, 2006, : 466 - +

← 1 2 3 4 5 →