Face and Gait Features for Person Identification Using Convolution Neural Network
Year: 2020
Type: article
Abstract: Automatic authentication of people has always been a challenging task, especially when it has to
deal with the large datasets and the robustness against the factors affecting recognition such as pose variation,
subject to camera angle, illumination, poor-quality data, and occlusion, etc. Hence deep learning proves to be a
great remedy to overcome the above problems. So, we have designed architecture to identify people by fusing
their gait and face biometric. Traits using multi hidden Layer Deep Convolution Neural Network (DCNN). In
our work, the concept of Gait Energy Images (GEIs) is used to represent the characteristics of human gait. The
GEIs and face sequences from the same individual are firstly resized into vectors of the same size after some
sort of pre-processing. After that, both the vectors are fused, and the output is fed to the DCNN for feature
extraction and classification. Our DCNN is self-possessed of triplets of Complication, Relu and Max-Pooling
Layers which is trailed by a fully connected Layer and a SoftMax Regression Layer. Classification is done at the
end to get the recognition results using a classification layer. The proposed DCNN model is tested upon the
publicly available CASIA Gait Dataset B and ORL Face Dataset and a recognition accuracy of 95% is achieved
on the test dataset.
Source: Journal of Critical Reviews
Authors: Namrata Chopra, Abhishek Thakur
Language: English
Subfield: Biomedical Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W3042402870
