Document Type

Journal Article

Publication Title

Data

Volume

8

Issue

10

Publisher

MDPI

School

School of Science

RAS ID

64552

Comments

Rahman, M. A., Aonty, S. S., Deb, K., & Sarker, I. H. (2023). Attention-based human age estimation from face images to enhance public security. Data, 8(10), article 145. https://doi.org/10.3390/data8100145

Abstract

Age estimation from facial images has gained significant attention due to its practical applications such as public security. However, one of the major challenges faced in this field is the limited availability of comprehensive training data. Moreover, due to the gradual nature of aging, similar-aged faces tend to share similarities despite their race, gender, or location. Recent studies on age estimation utilize convolutional neural networks (CNN), treating every facial region equally and disregarding potentially informative patches that contain age-specific details. Therefore, an attention module can be used to focus extra attention on important patches in the image. In this study, tests are conducted on different attention modules, namely CBAM, SENet, and Self-attention, implemented with a convolutional neural network. The focus is on developing a lightweight model that requires a low number of parameters. A merged dataset and other cutting-edge datasets are used to test the proposed model’s performance. In addition, transfer learning is used alongside the scratch CNN model to achieve optimal performance more efficiently. Experimental results on different aging face databases show the remarkable advantages of the proposed attention-based CNN model over the conventional CNN model by attaining the lowest mean absolute error and the lowest number of parameters with a better cumulative score.

DOI

10.3390/data8100145

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

 
COinS