Artificial intelligence

Abstract: This paper explores machine learning algorithms for face recognition, there pro and cons and investigate methods of their improvement. In paper is proposed usage of histogram of oriented gradients(HOG), L2-norm and face landmarks estimation for achieving most accurate result of face recognition.

Keywords: face recognition, machine learning, Histogram of oriented gradients, face landmarks estimation, Python, dblib

References:

1. Li S.Z. Handbook of Face Recognition / S.Z. Li, A. Jain; – М.: Springer New York, 2011. – 398 p.
2. Histograms of Oriented Gradients for Human Detection // [Електр. Ресурс]. – Режим доступу: https://lear.inrialpes.fr/people/triggs/pubs/Dalal-cvpr05.pdf
3. Theodoridis S. Pattern Recognition 4th Edition / S. Theodoridis, K. Koutroumbas; - М.: Academic Press, 2008. – 981 p.
4. FaceNet: A Unified Embedding for Face Recognition and Clustering // [Електр. Ресурс]. – Режим доступу: https://www.cv-foundation.org/openaccess/content_cvpr_2015 /app/1A_089.pdf
5. OpenFace. Free and open source face recognition with deep neural networks // [Електр. Ресурс]. – Режим доступу: https://cmusatyalab.github.io/openface/
6. Ma Yu. Support Vector Machines Applications / Yu. Ma, G. Guo; – M.: Springer Science & Business Media, 2014. – 302 p.
7. Apple Special Event // [Електр. Ресурс]. – Режим доступу: https://www.apple.com/apple-events/september- 2017/.
8. Face ID // [Електр. Ресурс]. – Режим доступу: https://en.wikipedia.org/wiki/Face_ID
9. DeepFace: Closing the Gap to Human-Level Performance in Face Verification // [Електр. Ресурс]. – Режим доступу: https://research.fb.com/publications/deepface-closing-the-gap-to-human-level-performance-in- face-verification/