<![CDATA[Gender identification is a crucial technique that can enhance the performance of authentication systems. Due to its variety of applications, human gender detection, a component of face recognition, has drawn a lot of interest. Previous studies on gender identification have relied on static features of the body, such as the face, eyebrows, hands, bodies, fingernails, etc. The abundance of face picture datasets available today has led to the development of several effective machine learning and deep learning techniques. When using classical machine learning techniques, it is essential to extract precise features from datasets in order to obtain favorable classification results. c Deep neural networks have the ability to explore hidden and unexpected feature sets, improving classification performance above conventional machine learning techniques. It can address the issue of the variable nature of facial signals between origins, which makes accurate feature extraction challenging. The effectiveness of the pre-trained DNN models is examined in this study when there is a dearth of data. Due to this issue, only the areas of the one eye picture with brows were considered in this study to classify the gender, as opposed to the entire face. The performance findings indicate that EfficientNetb7 is the best model and give better accuracy as compared to Xception, InceptionResNetV2, VGG16 and Resnet50.]]>
Copyrights © 2023
