Multi-level Multi-scale Deep Feature Encoding for Chronological Age Estimation from OPG Images

  • Sultan Alkaabi Mr
Keywords: Age estimation; DASNET; CNNs; ConvLSTM; Atrous pyramid convolution

Abstract

Age estimation is the first of the victims in forensic dentistry especially if the bodies have started to decompose. However, when the task involves Manually examining, the accuracy can decrease due varying experience of the experts, the results may also be varying from experts. To improve speed and accuracy of the age estimation process using forensic dentistry, researchers have proposed Convolutional Neural Network for Dental Age and Sex Network estimation (DASNET). However, pooling and scalar outputs of CNNs could not allow to get the equivariance due to the dental extraction complexity from panoramic images including jaws, teeth, lesions and carries. We developed a deep auto-encoder decoder architecture, which estimate the age based on semantic and structural feature representation. we propose Convolution Long Short Term Memory (ConvLSTM) to capture the correlation of features and generates high level representation of features. For the representation of the generated features, we utilize Atrous pyramid convolution to produce a multiscale representation. The propose techniques successfully reduces mean error to 0.75 years, as opposed to 0.93 years of the DASNET.

Published
2022-10-28
Section
Topic: Machine Learning Based Techniques for Image and Video Processing