Snack Food Texture Estimation by Neural Network

被引：1

作者：

Kato, Shigeru ^{[1
]}

Ito, Ryuji ^{[1
]}

Wada, Naoki ^{[1
]}

Kagawa, Tomomichi ^{[1
]}

Yamamoto, Masayoshi ^{[1
]}

机构：

[1] Natl Inst Technol, Niihama Coll, Niihama, Ehime, Japan

来源：

2018 JOINT 10TH INTERNATIONAL CONFERENCE ON SOFT COMPUTING AND INTELLIGENT SYSTEMS (SCIS) AND 19TH INTERNATIONAL SYMPOSIUM ON ADVANCED INTELLIGENT SYSTEMS (ISIS) | 2018年

关键词：

food texture; neural network; human sensibility; soft computing; CLASSIFICATION;

D O I：

10.1109/SCIS-ISIS.2018.00097

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 describes a system which estimates food texture of three kinds of snacks such as potato chips, wafers and cookies. The system consists of an original equipment and a neural network model. The equipment observes the load and the sound simultaneously when the food is probed. The system calculates input parameters of the neural network model. The parameters are characteristic values expressing the load change and the sound data. The model outputs numerical value ranged [0,1] which expresses the level of the textures such as "crunchiness" and "crispness". The teaching value of the neural network is determined by a sensory test by 30 subjects aged between 16 and 19 in order to reflect general human sensibility. In the experiment, it is found that the neural network model outputs moderate texture value of the snacks which are not used for training the model.

引用

页码：548 / 553

页数：6

共 15 条

[1]

[Anonymous], 2015, ARXIV PREPRINT ARXIV

[2] A review of acoustic research for studying the sensory perception of crisp, crunchy and crackly textures [J].