Approach to Object Hardness Prediction by Rubber Ball Hardness Prediction Using Capsule Network

Authors

Shota Shindo
Next Solutions Corporation Limited
Takaaki Goto
Toyo University, Department of Information Sciences and Arts
Kensei Tsuchida
Toyo University, Department of Information Sciences and Arts
DOI: https://doi.org/10.52549/ijeei.v9i4.3275
Download PDF
Published: 30-11-2021 • Issue: Vol. 9 No. 4: December 2021 • Section: Computer_&_Informatics

Abstract

A hardness is often used as an index to compare similar objects such as fruits or wood. To measure an object’s hardness, a hardness meter is required, and certain conditions must be met. The conditions are that the hardness meter is compatible with the object and must be close at hand. This research shows the possibility of measuring hardness without a hardness meter using a neural network. The method employs machine learning using a capsule network (CapsNet) of a neural network model. This research experimented using CapsNet with routing-by-agreement, CapsNet with expectation-maximization routing (EM routing) and the EM routing method with the addition of Tasks-Constrained Deep Convolutional Network (TCDCN). The four-layer CapsNet with EM routing implemented has achieved the state-of-the-art.  Multi-layered CapsNet with EM routing was a very effective method for regression analysis as well. And, CapsNet has higher discriminative power using EM-routing than routing-by-agreement.


Keywords: Capsule network; Hardness; Regression analysis; Image process; Neural network

Author Biographies

Shota Shindo

Shota Shindo received the master degree in computer science from the Department of Information Sciences and Arts , Toyo University, in 2020. His current research interests include machine learning, Computer learning theory, neural networks, Pattern recognition and artificial intelligence.


Takaaki Goto
Takaaki Goto graduated with a Doctor of Engineering from Toyo University in 2009. From 2009 to 2015, He had been a Project Assistant Professor at the University of Electro-Communications in Japan. He was an Associate Professor at Ryutsu Keizai University in Japan from 2015 to 2019. Now he is an Associate Professor at Toyo University, Japan. His main research interests are applications of graph grammars, visual languages, and software development environments. He is a member of ACM, IEEE, ACIS, ISCA, IEICE, and IPSJ.
Kensei Tsuchida
Kensei Tsuchida received M.S. and D.S. degrees in mathematics from Waseda University in 1984 and 1994, respectively. He was a member of the Software Engineering Development Laboratory, NEC Corporation in 1984–1990. From 1990 to 1992, he was a Research Associate of the Department of Industrial Engineering and Management at Kanagawa University. In 1992 he joined Toyo University, where he was an Instructor until 1995 and an associate professor from 1995 to 2002, and a Professor from 2002 to 2009 at the Department of Information and Computer Sciences, and since 2009 he has been a Professor of Faculty of Information Sciences and Arts. He was a Visiting Associate Professor of the Department of Computer Science at Oregon State University from 1997 to 1998. His research interests include software visualization, human interface, graph languages, and graph algorithms. He is a member of IPSJ, IEICE Japan and IEEE Computer Society.

How to Cite

[1]
S. Shindo, T. Goto, and K. Tsuchida, “Approach to Object Hardness Prediction by Rubber Ball Hardness Prediction Using Capsule Network”, IJEEI, vol. 9, no. 4, pp. 943–953, Nov. 2021, doi: 10.52549/ijeei.v9i4.3275.