Facial Recognition Using a Lightweight Deep Neural Networks
Keywords:Facial Recognition, Deep Learning, Convolutional Neural Network
Current facial recognition systems are still far away from the capability of the humanâ€™s face perception. Facial recognition systems can continue to be improved as technology evolves. The task of face recognition has been heavily explored in recent years. In this research, we provide our initial idea in developing Lightweight Deep Neural Networks for facial recognition. Although our goal was to create an optimal model that would exceed current facial recognition model performance, we could experiment and discover alternative approaches to multi-class facial recognition/classification. We tested with a dataset of 2800 images of men and women with specified image sizes. We created three CNN with various architectures, which we used to train with the chosen dataset for 20, 50, 100, and 200 classes per model. The experimental results exhibit the challenges of increasing the complexity of neural networks. From these results, we concluded that a Light CNN Model with a small number of layers had an average test accuracy of 94.19%, which was the best classification performance on unseen data.
- G. H. Bower and M. B. Karlin, â€œDepth of processing pictures of faces and recognition memory,â€ Journal of Experimental Psychology, vol. 103, pp. 751â€“757, 1974. [Online]. Available: https://doi.org/10.1037/h0037190
- H. Xu, X. Su, Y. Wang, H. Cai, K. Cui, and X. Chen, â€œAutomatic bridge crack detection using a convolutional neural network,â€ Applied Sciences, vol. 9, no. 14, p. 2867, 2019.
- E. Florio. (2018) At Marriott, you can now check in with your face. [Online]. Available: https://www.cntraveler.com/story/marriott-alibaba-facial-recognition-hotel-check-in
- C. Clifford. (2018) You can pay for your burger with your face at this fast-food restaurant, thanks to a.i. [Online]. Available: https://www.cnbc.com/2018/02/02/pay-with-facial-recognition-a-i-at-caliburger-in-pasadena-california.html
- G. Barkho. (2019) Walmart confirms the use of ai-powered cameras to detect stealing. [Online]. Available: https://observer.com/2019/06/walmart-ai-cameras-detect-stealing/
- Apple Press Release. (2017) The future is here: iphone x. [Online]. Available: https://www.apple.com/newsroom/2017/09/ the-future-is-here-iphone-x/
-  S. Oâ€™Dea. (2021) Share of smartphone users that use an Apple iPhone in the United States from 2014 to 2021. [Online]. Available: https://www.statista.com/statistics/236550/percentage-of-us-population-that-own-a-iphone-smartphone/
- J.-C. Chen, R. Ranjan, S. Sankaranarayanan, A. Kumar, C.-H. Chen, V. M. Patel, C. D. Castillo, and R. Chellappa, â€œUnconstrained still/video-based face verification with deep convolutional neural networks,â€ 2017.
- D. S. Abdelminaam, A. M. Almansori, M. Taha, and E. Badr, â€œA deep facial recognition system using computational intelligent algorithms,â€ PLOS ONE, vol. 15, no. 12.
- Grand View Research. (2020) Facial recognition market size, share trends analysis report by technology (2d, 3d), by application (emotion recognition, attendance tracking monitoring), by end-use, and segment forecasts, 2020 - 2027. [Online]. Available: https://www.grandviewresearch.com/industry-analysis/facial-recognition-market
- D. Castro and M. McLaughlin, â€œSurvey: Few Americans want government to limit use of facial recognition technology, particularly for public safety or airport screening,â€ 2019.
-  P. J. Grother, P. J. Grother, and M. Ngan, Face recognition vendor test (frvt). US Department of Commerce, National Institute of Standards and Technology, 2014.
- W. Crumpler, â€œHow accurate are facial recognition systemsâ€”and why does it matter,â€ Center for Strategic and International Studies, vol. 14, 2020.
- S. Saha, â€œA comprehensive guide to convolutional neural networks-the eli5 way,â€ Dec 2018. [Online]. Available: https://towardsdatascience.com/a-comprehensive-guide-to-convolutional-neural-networks-the-eli5-way-3bd2b1164a53
- W. You, C. Shen, X. Guo, X. Jiang, J. Shi, and Z. Zhu, â€œA hybrid technique based on convolutional neural network and support vector regression for intelligent diagnosis of rotating machinery,â€ Advances in Mechanical Engineering, vol. 9, no. 6, 2017.
- G. Guo and N. Zhang, â€œA survey on deep learning-based face recognition,â€ Computer Vision and Image Understanding, vol. 189, p. 102805, 2019.
- S. Almabdy and L. Elrefaei, â€œDeep convolutional neural network-based approaches for face recognition,â€ Applied Sciences, vol. 9, no. 20, 2019.
- M. Wang and W. Deng, â€œDeep face recognition: A survey,â€ Neurocomputing, vol. 429, p. 215â€“244, Mar 2021. [Online]. Available: http://dx.doi.org/10.1016/j.neucom.2020.10.081
- V. Sharma and P. C. says:, â€œRole of convolutional layer in convolutional neural networks,â€ Oct 2020. [Online]. Available: https://vinodsblog.com/2020/05/03/role-of-convolutional-layer-in-cnn/#:~:text=ConvolutionalLayerT1 textendashAnOutlook,i.e.extractfeaturesfromit.
- S. Khosla. (2019) Cnn | introduction to pooling layer. [Online]. Available: https://www.geeksforgeeks.org/cnn-introduction-to-pooling-layer/
- Deeplizard. (2017) Accuracy, precision, recall or f1? [Online]. Available: https://deeplizard.com/learn/video/m0pIlLfpXWE
- S. Kansal. (2017) A quick guide to activation functions in deep learning. [Online]. Available: https://towardsdatascience.com/a-quick-guide-to-activation-functions-in-deep-learning-4042e7addd5b
- Deshpande, â€œA beginnerâ€™s guide to understanding convolutional neural networks. 2016,â€ URl: adeshpande3. github.io/adeshpande3. github. io/A-Beginnerâ€™s-Guide-To-Understanding-Convolutional-Neural-Networks, 2020.
- K. P. Shung. (2018) Accuracy, precision, recall or f1? [Online]. Available: https://towardsdatascience.com/accuracy-precision-recall-or-f1-331fb37c5cb9
- W. Koehrsen. (2018) Beyond accuracy: Precision and recall. [Online]. Available: https://towardsdatascience.com/beyond-accuracy-precision-and-recall-3da06bea9f6c
- C. E. Thomaz, â€œFei face database.â€ [Online]. Available: https://fei.edu.br/~cet/facedatabase.html