Publication:
Enhancing CNN for Forensics Age Estimation Using CGAN and Pseudo-Labelling

Simple item page
dc.citedby3
dc.contributor.authorAlkaabi S.en_US
dc.contributor.authorYussof S.en_US
dc.contributor.authorAl-Mulla S.en_US
dc.contributor.authorid59070935100en_US
dc.contributor.authorid16023225600en_US
dc.contributor.authorid36473139200en_US
dc.date.accessioned2024-10-14T03:21:51Z
dc.date.available2024-10-14T03:21:51Z
dc.date.issued2023
dc.description.abstractAge estimation using forensics odontology is an important process in identifying victims in criminal or mass disaster cases. Traditionally, this process is done manually by human expert. However, the speed and accuracy may vary depending on the expertise level of the human expert and other human factors such as level of fatigue and attentiveness. To improve the recognition speed and consistency, researchers have proposed automated age estimation using deep learning techniques such as Convolutional Neural Network (CNN). CNN requires many training images to obtain high percentage of recognition accuracy. Unfortunately, it is very difficult to get large number of samples of dental images for training the CNN due to the need to comply to privacy acts. A promising solution to this problem is a technique called Generative Adversarial Network (GAN). GAN is a technique that can generate synthetic images that has similar statistics as the training set. A variation of GAN called Conditional GAN (CGAN) enables the generation of the synthetic images to be controlled more precisely such that only the specified type of images will be generated. This paper proposes a CGAN for generating new dental images to increase the number of images available for training a CNN model to perform age estimation. We also propose a pseudo-labelling technique to label the generated images with proper age and gender. We used the combination of real and generated images to train Dental Age and Sex Net (DASNET), which is a CNN model for dental age estimation. Based on the experiment conducted, the accuracy, coefficient of determination (R2) and Absolute Error (AE) of DASNET have improved to 87%, 0.85 and 1.18 years respectively as opposed to 74%, 0.72 and 3.45 years when DASNET is trained using real, but smaller number of images. � 2023 Tech Science Press. All rights reserved.en_US
dc.description.natureFinalen_US
dc.identifier.doi10.32604/cmc.2023.029914
dc.identifier.epage2516
dc.identifier.issue2
dc.identifier.scopus2-s2.0-85141892582
dc.identifier.spage2499
dc.identifier.urihttps://www.scopus.com/inward/record.uri?eid=2-s2.0-85141892582&doi=10.32604%2fcmc.2023.029914&partnerID=40&md5=3d6e4208eda7c0816b49d401c1b39264
dc.identifier.urihttps://irepository.uniten.edu.my/handle/123456789/34700
dc.identifier.volume74
dc.pagecount17
dc.publisherTech Science Pressen_US
dc.relation.ispartofAll Open Access
dc.relation.ispartofGold Open Access
dc.sourceScopus
dc.sourcetitleComputers, Materials and Continua
dc.subjectage estimation
dc.subjectconvolutional neural network
dc.subjectDental forensics
dc.subjectgenerative adversarial network
dc.subjectpseudo-labelling
dc.subjectConvolution
dc.subjectConvolutional neural networks
dc.subjectDeep learning
dc.subjectDigital forensics
dc.subjectImage enhancement
dc.subjectNeural network models
dc.subjectAge estimation
dc.subjectConvolutional neural network
dc.subjectDental forensic
dc.subjectDental images
dc.subjectForensic odontology
dc.subjectHuman expert
dc.subjectLabelings
dc.subjectNeural network model
dc.subjectPseudo-labeling
dc.subjectSynthetic images
dc.subjectGenerative adversarial networks
dc.titleEnhancing CNN for Forensics Age Estimation Using CGAN and Pseudo-Labellingen_US
dc.typeArticleen_US
dspace.entity.typePublication
Files
Collections
SCOPUS