DSpace Repository

Customer Face Analysis for Retail Analytics

Show simple item record

dc.contributor.advisor Dailey, Dr. Matthew N.
dc.contributor.author Karim, Nabil Tahmidul
dc.contributor.other Ekpanyapong, Dr. Mongkol
dc.contributor.other Esichaikul, Dr. Vatcharaporn
dc.date.accessioned 2017-05-08T03:08:39Z
dc.date.available 2017-05-08T03:08:39Z
dc.date.issued 2017-05
dc.identifier.other AIT
dc.identifier.other AIT
dc.identifier.uri http://www.cs.ait.ac.th/xmlui/handle/123456789/853
dc.description.abstract Customer experience, feedback and keeping a customer engaging surrounding are influential factors for any retail store. Acknowledging these facts enhance the development of businesses. Acquiring constructive statistics about customer interests can be extremely valuable for retail stores. This thesis aims to provide such information through an automated system using deep learning methodology. In this thesis, I present an automated system where face detection and tracking are done to retrieve faces of customers coming to retail stores. Afterwards, Siamese convolutional neural network is used to determine if its an old or a new customer. Then I use the same face image to identify the person’s age and gender through two more convolutional neural networks. At every frame, another convolutional neural network is used to identify the person’s sentiment and record it in the database. A series of individual experiments show that each of the network achieve satisfactory results. This prototype can provide a base for further research into using face features for retail analytics. en_US
dc.description.sponsorship AIT Fellowship en_US
dc.language.iso en_US en_US
dc.publisher AIT en_US
dc.subject Face Verification, Repeat-customer Analaysis, Age Classification, Gender Classification, Emotion Classification, Convolutional Neural Network, Siamese Convolutional Neural Network, Caffe. en_US
dc.title Customer Face Analysis for Retail Analytics en_US
dc.type Thesis en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search DSpace


Advanced Search

Browse

My Account