Treffer: Gender, age, and ethnicity estimation by image processing.

Today, with the increasing interest in technology, very useful studies are carried out in the field of image processing. Image technologies are also used in many fields such as security, defense, medicine, and industry. In this study, age, gender, and ethnicity were determined in the images by employing various deep-learning techniques and constructing a custom model using Convolutional Neural Networks (CNN). The dataset, consisting of 23,705 images obtained from the Kaggle dataset named "Face Data," was utilized for the analysis. The images were categorized based on gender, race, and age within the application, and the accuracy and losses of the results were visualized through graphs. Moreover, an interface was created using the Python Flask library, enabling real-time analysis of images captured from the camera to determine age, gender, and race. Among the 23,705 images, approximately 12,000 were male profiles and 11,000 were female profiles. These profiles were further classified into 5 distinct ethnicities as specified in the dataset. The ethnicities in the application were represented as follows: 0 for White, 1 for Black, 2 for Asian, 3 for Indian, and 4 for others. The most challenging aspect of this study is the variability of images due to factors such as posture, pose angle, brightness, and resolution at the time of shooting. Despite these challenges, the developed models showcased promising results, as evidenced by the accuracy metrics and visual representations provided in the study. The integration of real-time image analysis through the Python Flask interface enhances the practical applicability of the proposed techniques in various scenarios. [ABSTRACT FROM AUTHOR]

Copyright of Dicle University Journal of Engineering / Dicle Üniversitesi Mühendislik Dergisi is the property of Dicle Universitesi and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)