Görüntü İşleme ile Yaş ve Cinsiyet Tahmini

Abstract

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 found in the image by using different deep learning techniques and by building our own model in CNN. The 23705 images taken from the csv file named Face Data taken from Kaggle were categorized as different gender, race, and age within the application and the accuracy and losses of the results were transferred with graphs. In addition, by creating an interface with the help of the Python flask library, the results of the snapshot taken from the camera (age, gender, and race) can also be found. Out of the 23705 images, approximately 12000 male and 11000 female profiles were obtained. These profiles were classified according to 5 different genetics specified in the dataset. The genetics in the application (0 represented White, 1 represented Black, 2 represented Asian, 3 represented Indian, 4 represented Others.) The most difficult part of this study is that the picture changes depending on many factors such as posture, pose angle, brightness, and resolution at the time of shooting..

Keywords

• cnn
• image processing
• python
• open cv
• classification

Gender and Age Estimation By Image Processing

Abstract

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 found in the image by using different deep learning techniques and by building our own model in CNN. The 23705 images taken from the csv file named Face Data taken from Kaggle were categorized as different gender, race, and age within the application and the accuracy and losses of the results were transferred with graphs. In addition, by creating an interface with the help of the Python flask library, the results of the snapshot taken from the camera (age, gender, and race) can also be found. Out of the 23705 images, approximately 12000 male and 11000 female profiles were obtained. These profiles were classified according to 5 different genetics specified in the dataset. The genetics in the application (0 represented White, 1 represented Black, 2 represented Asian, 3 represented Indian, 4 represented Others.) The most difficult part of this study is that the picture changes depending on many factors such as posture, pose angle, brightness, and resolution at the time of shooting..

Keywords

• CNN
• İmage processing
• python
• open cv
• classification

References

1. [1] Ç. Kılınç, “Why Do We Use Matrices for Image Processing?” [Online]. Available: https://medium.com/@cgtyklnc/why-do-we-use-matrices-for-image-processing-3b24a59abe4f/, Accessed on: Jan. 6, 2023.
2. [2] N. Arora, "Age, Gender, and Ethnicity Face Data." [Online]. Available: https://www.kaggle.com/datasets/nipunarora8/age-gender-and-ethnicity-face-data-csv/, Accessed on: Sep. 2, 2023.
3. [3] Protopars, “Derin öğrenme (deep learning) nedir?” [Online]. Available: https://www.protopars.com/derin-ogrenme-deep-learning-nedir/, Accessed on: May. 13, 2023.
4. [4] Statology, “The easiest way to use seaborn: import seaborn as sns”. [Online]. Available: https://www.statology.org/import-seaborn-as-sns/, Accessed on: May. 19, 2023.
5. [5] T. Ergin, “Keras ile derin öğrenme model oluşturma” [Online]. Available: https://medium.com/@tuncerergin-/keras-ile-derin-ogrenme-model-olusturma-4b4ffdc35323, Oct 2, 2018. Accessed on: June 11, 2023.
6. [6] E. Uzun,“Makine öğrenmesi” [Online]. Available: https://erdincuzun.com/makine_ogrenmesi/makine-ogrenmesi-metotlari/, Accessed on: Jan. 10, 2023.
7. [7] M. F. Akca, “Sınıflandırma problemlerindeki metrikler”. [Online]. Available: https://medium.com/deep-learning-turkiye/s%C4%B1n%C4%B1fland%C4%B1rma-problemlerindeki-metrikler-33ee5f30f8eb, Accessed on: April. 15, 2023.
8. [8] M. U. Uçar, “Recognition of students in the classroom environment and detection of distractions with real-time image processing.” Master's thesis, Department of Electrical and Electronics Engineering, İskenderun Technical University, Hatay, 2019.

9. [9] A. Günay, and V. Nabiyev, “Investigating the effects of facial regions to age estimation.” Türkiye Bilişim Vakfı Journal of Computer Science and Engineering, vol. 9, no. 2, pp.1-10, 2016.
10. [10] F. Ayata, and H. Çavuş, “Performance tests of ESA, YGH-DVM and DSA algorithms used in face recognition systems.” Fırat University Journal of Science and Technology, vol. 34, no.1, pp. 39-48, 2022.
11. [11] A. Eldem, H. Eldem, and A. Palalı, “Face detection system development with image processing techniques”. Bitlis Eren University Journal of Science and Technology, vol.6, no.2, pp. 44-48, 2017.
12. [12] G. Gündüz, and İ. H. Cedimoğlu, “Gender estimation with image by using deep learning algorithms.” Sakarya University Journal of Computer and Information Sciences, vol.2, no.1, pp. 9-17, 2019.
13. [13] K. Kayaalp, and S. Metlek, “Detection of fish species with deep learning.” International Journal of 3D Printing Technologies and Digital Industry, vol.5, no.3, pp. 569-576. 2021.
14. [14] Ö. Toprak, “Age estimation with image processing techniques,” Master's thesis, Institute of Science and Technology, Maltepe University, Istanbul, 2019.