Research on the visually impaired individuals shopping with artificial intelligence image recognition assistance

Chia Hui Feng, Ju Yen Hsieh, Yu Hsiu Hung, Chung Jen Chen, Cheng Hung Chen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Shopping is an indispensable part of daily life. It is an easy task for people with healthy eyes. However, it remains a big problem for the visually impaired individuals. Today, the visually impaired individuals have to be accompanied by their family or guided by the store escort when shopping. It is difficult for them to shop alone. This research develops an artificial intelligence image recognition auxiliary device utilizing the artificial intelligence technology Convolutional Neural Network (CNN), providing smart image recognition modules to assist the visually impaired individuals while shopping. CNN is the most effective deep learning algorithm in the field of machine vision, its ability to compare details of product exterior features makes product recognition more efficient via accurate model training. This study experiments task-oriented shopping in three shopping models, (1) self-shopping, (2) accompanied shopping, (3) device assisted shopping. It measures through three indicators: shopping time, accuracy in choosing the correct product, and device satisfaction. The research subjects are 18 college students, 8 male students and 10 female students. The subjects are blindfolded, simulating the visually impaired individuals to perform experiments in a state without any vision. one-way repeated-measures ANOVA is used to explore the differences among the three shopping models. Surveys are collected at the end of the experiment to analyze the degrees of satisfaction for the AIoT device. The results of this study are: (1) task operation time of the three shopping models are significantly different p =.000, and gender difference has no significant impact. (2) the task operation accuracy rate of the three shopping models are significantly different p =.000, gender difference also has significant impact p =.000. The accuracy rate for self-shopping is 39%, 12.5% for men and 60% for women. The accuracy rate for companied shopping is 97.25%, 93.75% for men and 100% for women. The accuracy rate for device assisted shopping is 90.25%, 87.5% for men and 92.5% for women. (3) highest score in satisfactory rating is the extensiveness of product information audio at an average at 4.5. Satisfactory rating of product information audio accurateness averages at 4.44 points. As to device effective for shopping assistance, the average satisfactory rating is 4.17. And the average satisfactory rating for device operation usability is 4.00.

Original languageEnglish
Title of host publicationUniversal Access in Human-Computer Interaction. Applications and Practice - 14th International Conference, UAHCI 2020, Held as Part of the 22nd HCI International Conference, HCII 2020, Proceedings
EditorsMargherita Antona, Constantine Stephanidis
PublisherSpringer
Pages518-531
Number of pages14
ISBN (Print)9783030491079
DOIs
Publication statusPublished - 2020
Event14th International Conference on Universal Access in Human-Computer Interaction, UAHCI 2020, held as part of the 22nd International Conference on Human-Computer Interaction, HCII 2020 - Copenhagen, Denmark
Duration: 2020 Jul 192020 Jul 24

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12189 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference14th International Conference on Universal Access in Human-Computer Interaction, UAHCI 2020, held as part of the 22nd International Conference on Human-Computer Interaction, HCII 2020
Country/TerritoryDenmark
CityCopenhagen
Period20-07-1920-07-24

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Research on the visually impaired individuals shopping with artificial intelligence image recognition assistance'. Together they form a unique fingerprint.

Cite this