Feature Selection From MagFace Face Recognition Model With Optimization Algorithms

Mehmet Fatih Ozdemır; Davut Hanbay

Makale detayları

Feature Selection From MagFace Face Recognition Model With Optimization Algorithms

Yazarlar:

Mehmet Fatih Ozdemır | Davut Hanbay

Yayıncı:

Fırat University

Ana Başlıklar:

Derin Öğrenme

Dergi:

Fırat Üniversitesi Mühendislik Bilimleri Dergisi Vol.35 Issue 2, pp.561-567

Belge Tipi:

Makale

Yayın Tarihi:

1/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001287530

pISSN:

1308-9072

Anahtar kelimeler:

Özet:

 In recent years, many studies have been carried out in the field of artificial intelligence in the literature with the development of equipment. Face recognition algorithms have an important place among these developments. Among the face recognition algorithms, the most successful ones are usually deep learning approaches. Models such as SphereFace, CosFace, ArcFace, and MagFace are important deep learning models in the literature. Despite their success, deep learning models are often computationally costly. Therefore, advanced methods are needed to reduce the computational load for these models. One of the most valid methods for this is to choose the most valuable one among embedding features for face recognition. Thus, cost can be reduced, and accuracy values can be increased even more. In this study, the most valuable of the 512 embedded features in the MagFace model was tried to be obtained by using PSO, GA, SCA, and DE optimization algorithms. As a result, accuracy values of 99.83, 98.57, and 98.65 were reached for 193, 252, and 280 features selected in the LFW, CFP, and AGEDB datasets, respectively.

 In recent years, many studies have been carried out in the field of artificial intelligence in the literature with the development of equipment. Face recognition algorithms have an important place among these developments. Among the face recognition algorithms, the most successful ones are usually deep learning approaches. Models such as SphereFace, CosFace, ArcFace, and MagFace are important deep learning models in the literature. Despite their success, deep learning models are often computationally costly. Therefore, advanced methods are needed to reduce the computational load for these models. One of the most valid methods for this is to choose the most valuable one among embedding features for face recognition. Thus, cost can be reduced, and accuracy values can be increased even more. In this study, the most valuable of the 512 embedded features in the MagFace model was tried to be obtained by using PSO, GA, SCA, and DE optimization algorithms. As a result, accuracy values of 99.83, 98.57, and 98.65 were reached for 193, 252, and 280 features selected in the LFW, CFP, and AGEDB datasets, respectively.

 In recent years, many studies have been carried out in the field of artificial intelligence in the literature with the development of equipment. Face recognition algorithms have an important place among these developments. Among the face recognition algorithms, the most successful ones are usually deep learning approaches. Models such as SphereFace, CosFace, ArcFace, and MagFace are important deep learning models in the literature. Despite their success, deep learning models are often computationally costly. Therefore, advanced methods are needed to reduce the computational load for these models. One of the most valid methods for this is to choose the most valuable one among embedding features for face recognition. Thus, cost can be reduced, and accuracy values can be increased even more. In this study, the most valuable of the 512 embedded features in the MagFace model was tried to be obtained by using PSO, GA, SCA, and DE optimization algorithms. As a result, accuracy values of 99.83, 98.57, and 98.65 were reached for 193, 252, and 280 features selected in the LFW, CFP, and AGEDB datasets, respectively.

 Son yıllarda gelişen donanımlarla birlikte literatürde yapay zeka alanında birçok çalışma yapılmaktadır. Bu gelişmeler arasında yüz tanıma algoritmaları önemli bir yere sahiptir. Yüz tanıma algoritmaları arasında ise en başarılı olanları genellikle derin öğrenme yaklaşımlarıdır. SphereFace, CosFace, ArcFace, MagFace gibi modeller literatürde yer alan önemli derin öğrenme modelleridir. Derin öğrenme modelleri başarılarının aksine genellikle hesaplama açısından maliyetlidir. Bu nedenle, bu modeller için hesaplama yükünü azaltacak gelişmiş yöntemlere ihtiyaç duyulmaktadır. Bunun için en geçerli yöntemlerden biri gömülü yüz öznitelikleri arasından en değerli olanı seçmektir. Böylece maliyet düşürülebilir hatta başarı değerleri daha da arttırılabilir. Bu çalışmada PSO, GA, SCA, DE optimizasyon algoritmaları kullanılarak MagFace 512 gömülü özelliklerinin en değerlileri elde edilmeye çalışılmıştır. Sonuç olarak LFW, CFP, AGEDB veri setlerinde seçilen değerli 193, 252, 280 öznitelikleri sırasıyla 99.83, 98.57, 98.65 doğruluk değerlerine ulaşılmıştır.

Makale detayları

Feature Selection From MagFace Face Recognition Model With Optimization Algorithms

Yazarlar:

Mehmet Fatih Ozdemır | Davut Hanbay

Yayıncı:

Fırat University

Ana Başlıklar:

Derin Öğrenme

Dergi:

Fırat Üniversitesi Mühendislik Bilimleri Dergisi Vol.35 Issue 2, pp.561-567

Belge Tipi:

Makale

Yayın Tarihi:

1/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001287530

pISSN:

1308-9072

Face Recognition

Feature Selection

Optimization

Yüz Tanıma

Öznitelik Seçimi

Optimizasyon

Sayı

Makale detayları

Feature Selection From MagFace Face Recognition Model With Optimization Algorithms

Yazarlar: Mehmet Fatih Ozdemır | Davut Hanbay

Yayıncı: Fırat University

Ana Başlıklar: Derin Öğrenme

Dergi: Fırat Üniversitesi Mühendislik Bilimleri Dergisi Vol.35 Issue 2, pp.561-567

Belge Tipi: Makale

Yayın Tarihi: 1/2023

Belge Dili: İngilizce

Document ID (Org No.): 2001287530

pISSN: 1308-9072

Face Recognition

Feature Selection

Optimization

Yüz Tanıma

Öznitelik Seçimi

Optimizasyon

Sayı

Yazarlar:

Mehmet Fatih Ozdemır | Davut Hanbay

Yayıncı:

Fırat University

Ana Başlıklar:

Derin Öğrenme

Dergi:

Fırat Üniversitesi Mühendislik Bilimleri Dergisi Vol.35 Issue 2, pp.561-567

Belge Tipi:

Makale

Yayın Tarihi:

1/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001287530

pISSN:

1308-9072