因為COVID-19 疫情大流行，人們在日常生活中都戴上了口罩。這事實影響了傳統不戴口罩人臉辨識方法的有效性。目前，大多數基於人工智慧的人臉辨識方法都需要大量的影像樣本。而且，在執行這些影像的標記步驟時將會耗費不少人力和時間。為了解決上述問題，在本論文中，我們提出一種用於有限樣本戴口罩人臉辨識的改進型分類器。我們的方法使用被稱為原型分類器(prototype classifier)的少樣本學習(few-shot learning, FSL)模型來實現分類算法。原型分類器執行距離計算過程以對身份進行分類。我們通過整合類別共變異數矩陣(class-covariance matrix)來改進距離計算。這使得該算法更適合複雜和遮擋的面部特徵。除此之外，我們使用銳度感知最小化(sharpness-aware minimization, SAM)。其被設計用來最小化損失值，同時了解損失的銳度，使識別更加準確。基於各種數據集之廣泛且深入的實驗顯示，我們的方法達到了顯著的結果，準確率值高達 95.3%。本論文所提出的方法所獲得的準確率高於比較基準 3.4%。簡而言之，在嚴峻的COVID-19疫情大流行情況下，我們的研究為有限身份樣本戴口罩人臉辨識之問題提供了可靠而且有效的解決方案。

Due to the COVID-19 pandemic, people wear masks in their daily life. This fact affects the effectiveness of traditional face recognition methods without wearing masks. At present, most face recognition methods based on artificial intelligence require a large number of image samples. Moreover, it will consume a lot of manpower and take a lot of time when performing the labeling steps for these images. In order to solve the above problems, in this thesis, we propose an improved classifier for masked face recognition with limited samples. Our method uses a few-shot learning (FSL) model, called a prototype classifier, to implement a classification algorithm. The prototype classifier performs a distance calculation process to classify identities. We improve the distance calculation by integrating the class-covariance matrix. The class-covariance matrix makes the algorithm more suitable for dealing with complex and occluded facial features. In addition to that, we use sharpness-aware minimization (SAM). It is designed to minimize the loss value while simultaneously being aware of its sharpness to make the recognition even more accurate. Extensive and in-depth experiments based on various datasets have shown that our method yields remarkable results, with accuracy values reaching 95.3%. The accuracy obtained by the proposed method in this thesis is 3.4% higher than the benchmark used for comparison. In short, in the critical COVID-19 pandemic situation, our study presents a reliable and efficient solution to the problem of limited sample per identity in masked face recognition.

[1] S. Minaee, A. Abdolrashidi, H. Su, M. Bennamoun, and D. Zhang, “Biometrics recognition using deep learning: A survey,” Artificial Intelligence Review, vol. 56, no. 1, pp. 1155–1187, January 2023.
[2] A. Alzu’bi, F. Albalas, T. Al-Hadhrami, L. B. Younis, and A. Bashayreh, “Masked face recognition using deep learning: A review,” Electronics, vol. 10, no. 21, pp. 1–35, October 2021.
[3] M. Wang and W. Deng, “Deep face recognition: A survey,” Neurocomputing, vol. 429, pp. 215–244, March 2021.
[4] C. Peng, X. Gao, N. Wang, and J. Li, “Sparse graphical representation based discriminant analysis for heterogeneous face recognition,” Signal Processing, vol. 156, pp. 46–61, March 2019.
[5] Y. Li, K. Guo, Y. Lu, and L. Liu, “Cropping and attention based approach for masked face recognition,” Applied Intelligence, vol. 51, no. 5, pp. 3012–3025, May 2021.
[6] H. N. Vu, M. H. Nguyen, and C. Pham, “Masked face recognition with convolutional neural networks and local binary patterns,” Applied Intelligence, vol. 52, no. 5, pp. 5497–5512, March 2022.
[7] E. Freud, D. Di Giammarino, A. Stajduhar, R. S. Rosenbaum, G. Avidan, and T. Ganel, “Recognition of masked faces in the era of the pandemic: No improvement despite extensive natural exposure,” Psychological Science, vol. 33, no. 10, pp. 1635–1650, October 2022.
[8] H. Du, H. Shi, Y. Liu, D. Zeng, and T. Mei, “Towards NIR-VIS masked face recognition,” arXiv:2104.06761, pp. 1–5, April 2021.
[9] J. Deng, J. Guo, X. An, Z. Zhu, and S. Zafeiriou, “Masked face recognition challenge: The insightface track report,” in Proc. IEEE International Conference on Computer Vision Workshops, Montreal, Canada, October 11–17, 2021, pp. 1437–1444.
[10] W. Moungsouy, T. Tawanbunjerd, N. Liamsomboon, and W. Kusakunniran, “Face recognition under mask-wearing based on residual inception networks,” Applied Computing and Informatics, vol. 18, no. 2, pp. 1–14, April 2022.
[11] L. Cao, X. Huo, Y. Guo, and K. Du, “Sketch face recognition via cascaded transformation generation network,” IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, vol. E104.A, no. 10, pp. 1403–1415, October 2021.
[12] B. Wang, J. Zheng, and C. L. P. Chen, “A survey on masked facial detection methods and datasets for fighting against COVID-19,” IEEE Transactions on Artificial Intelligence, vol. 3, no. 3, pp. 323–343, January 2022.
[13] R. V. Petrescu, “Face recognition as a biometric application,” SSRN Electronic Journal, vol. 3, no. 1, pp. 237–257, May 2019.
[14] A. Anwar and A. Raychowdhury, “Masked face recognition for secure authentication,” arXiv:2008.11104, pp. 1–8, August 2020.
[15] B. Mandal, A. Okeukwu, and Y. Theis, “Masked face recognition using ResNet-50,” arXiv:2104.08997, pp. 1–8, April 2021.
[16] W. Hariri, “Efficient masked face recognition method during the COVID-19 pandemic,” Signal, Image and Video Processing, vol. 16, no. 3, pp. 605–612, April 2022.
[17] G. Hinton, “Deep neural networks for acoustic modeling in speech recognition: The shared views of four research groups,” IEEE Signal Processing Magazine, vol. 29, no. 6, pp. 82–97, November 2012.
[18] X. Tan, S. Chen, Z.-H. Zhou, and F. Zhang, “Face recognition from a single image per person: A survey,” Pattern Recognition, vol. 39, no. 9, pp. 1725–1745, September 2006.
[19] S. Kadam and V. Vaidya, “Review and analysis of zero, one and few shot learning approaches,” in Proc. International Conference on Intelligent Systems Design and Applications, Vellore, India, December 6–18, 2020, pp. 100–112.
[20] D. Zeng, R. Veldhuis, and L. Spreeuwers, “A survey of face recognition techniques under occlusion,” IET Biometrics, vol. 10, no. 6, pp. 581–606, November 2021.
[21] Q. Zhu, Q. Mao, H. Jia, O. E. N. Noi, and J. Tu, “Convolutional relation network for facial expression recognition in the wild with few-shot learning,” Expert Systems with Applications, vol. 189, pp. 116–146, March 2022.
[22] Y. Song, T. Wang, P. Cai, S. K. Mondal, and J. P. Sahoo, “A comprehensive survey of few-shot learning: Evolution, applications, challenges, and opportunities,” ACM Computer Survey, pp. 1–24, February 2023.
[23] E. Schwartz, L. Karlinsky, R. Feris, R. Giryes, and A. Bronstein, “Baby steps towards few-shot learning with multiple semantics,” Pattern Recognition Letters, vol. 160, pp. 142–147, August 2022.
[24] H. Zhang, P. Koniusz, S. Jian, H. Li, and P. H. S. Torr, “Rethinking class relations: absolute-relative supervised and unsupervised few-shot learning,” in Proc. IEEE Conference on Computer Vision and Pattern Recognition, Nashville, USA, June 20–25, 2021, pp. 9427–9436.
[25] M. Fink, “Object classification from a single example utilizing class relevance metrics,” in Proc. Advances in Neural Information Processing Systems, Vancouver, Canada, July 1, 2004.
[26] Y. Yang, Y. Li, R. Zhang, J. Wang, and Z. Miao, “Robust compare network for few-shot learning,” IEEE Access, vol. 8, pp. 137966–137974, July 2020.
[27] Y. Zou, S. Zhang, K. Chen, Y. Tian, Y. Wang, and J. M. F. Moura, “Compositional few-shot recognition with primitive discovery and enhancing,” in Proc. ACM International Conference on Multimedia, New York, USA, October 12–16, 2020, pp. 156–164.
[28] R. Duan, D. Li, Q. Tong, T. Yang, X. Liu, and X. Liu, “A survey of few-shot learning: An effective method for intrusion detection,” Security and Communication Networks, vol. 2021, pp. 1–10, October 2021.
[29] P. Sirinam, N. Mathews, M. S. Rahman, and M. Wright, “Triplet fingerprinting: more practical and portable website fingerprinting with n-shot learning,” in Proc. ACM SIGSAC Conference on Computer and Communications Security, New York, USA, November 11–15, 2019, pp. 1131–1148.
[30] J. Kim and S. Chi, “A few-shot learning approach for database-free vision-based monitoring on construction sites,” Automation in Construction, vol. 124, pp. 103–116, April 2021.
[31] O. Vinyals, C. Blundell, T. Lillicrap, K. Kavukcuoglu, and D. Wierstra, “Matching networks for one shot learning,” arXiv:1606.04080, pp. 1–12, June 2016.
[32] L. M. Tassis and R. A. Krohling, “Few-shot learning for biotic stress classification of coffee leaves,” Artificial Intelligence in Agriculture, vol. 6, pp. 55–67, April 2022.
[33] Z. Zhan, J. Zhou, and B. Xu, “Fabric defect classification using prototypical network of few-shot learning algorithm,” Computers in Industry, vol. 138, pp. 103–118, June 2022.
[34] L. Chen, X. Tian, G. Chai, X. Zhang, and E. Chen, “A new CBAM-P-Net model for few-shot forest species classification using airborne hyperspectral images,” Remote Sensing, vol. 13, no. 7, pp. 126–139, March 2021.
[35] L. Bertinetto, J. F. Henriques, J. Valmadre, P. H. S. Torr, and A. Vedaldi, “Learning feed-forward one-shot learners,” arXiv:1606.05233, pp. 1–12, June 2016.
[36] M. Köhler, M. Eisenbach, and H.-M. Gross, “Few-shot object detection: A comprehensive survey,” arXiv:2112.11699, pp. 1–27, December 2021.
[37] J. Zhou, “Real-time visual object tracking via few-shot learning,” arXiv:2103.10130, pp. 1–11, March 2021.
[38] A. Brock, T. Lim, J. M. Ritchie, and N. Weston, “SMASH: one-shot model architecture search through hypernetworks,” arXiv:1708.05344, pp. 1–21, August 2017.
[39] T. Elsken, B. Staffler, J. H. Metzen, and F. Hutter, “Meta-learning of neural architectures for few-shot learning,” arXiv:1911.11090, pp. 1–15, November 2019.
[40] Y. Zhao, L. Wang, Y. Tian, R. Fonseca, and T. Guo, “Few-shot neural architecture search,” arXiv:2006.06863, pp. 1–14, June 2020.
[41] J. Snell, K. Swersky, and R. S. Zemel, “Prototypical networks for few-shot learning,” arXiv:1703.05175, pp. 1–13, March 2017.
[42] W.-Y. Chen, Y.-C. Liu, Z. Kira, Y.-C. F. Wang, and J.-B. Huang, “A closer look at few-shot classification,” arXiv:1904.04232, pp. 1–17, April 2019.
[43] M. Li, B. Huang, and G. Tian, “A comprehensive survey on 3D face recognition methods,” Engineering Applications of Artificial Intelligence, vol. 110, pp 104–119, April 2022.
[44] P. Weller, F. Aziz, S. Abdulatif, U. Schneider, and M. F. Huber, “A MIMO radar-based few-shot learning approach for human-id,” in Proc. European Signal Processing Conference, Belgrade, Serbia, August 29–September 2, 2022, pp. 1796–1800.
[45] L. Li, X. Mu, S. Li, and H. Peng, “A review of face recognition technology,” IEEE Access, vol. 8, pp. 139110–139120, July 2020.
[46] R. A. Waelen, “The struggle for recognition in the age of facial recognition technology,” AI and Ethics, vol. 3, no. 1, pp. 215–222, February 2023.
[47] Y. Wei, “Review of face recognition algorithms,” in Proc. International Conference on Biomedical Imaging, Signal Processing, Kitakyushu, Japan, September 27–29, 2020, pp. 28–31.
[48] R. Szeliski, Computer Vision, Springer, London, 2011.
[49] P. F. Cabana, “Technical and legal challenges of the use of automated facial recognition technologies for law enforcement and forensic purposes,” Artificial Intelligence, Social Harms and Human Rights, edited by A. Završnik and K. Simončič, pp. 35–54, January 2023.
[50] M. K. Rusia and D. K. Singh, “A comprehensive survey on techniques to handle face identity threats: challenges and opportunities,” Multimedia Tools and Applications, vol. 82, no. 2, pp. 1669–1748, January 2023.
[51] I. Khandokar, M. Hasan, F. Ernawan, S. Islam, and M. N. Kabir, “Handwritten character recognition using convolutional neural network,” Journal of Physics: Conference Series, vol. 1918, no. 4, pp 142–152, June 2021.
[52] Y. Luo, Y. Sun, and X. Bi, “Multiple attentional aggregation network for handwritten dongba character recognition,” Expert Systems with Applications, vol. 213, pp. 118–135, March 2023.
[53] M. C. G. Neri, O. O. V. Villegas, V. G. C. Sánchez, H. D. J. O. Domínguez, M. Nandayapa, and J. H. S. Azuela, “A methodology for character recognition and revision of the linear equations solving procedure,” Information Processing and Management, vol. 60, no. 1, pp. 103–118, January 2023.
[54] J. Gan, Y. Chen, B. Hu, J. Leng, W. Wang, and X. Gao, “Characters as graphs: interpretable handwritten chinese character recognition via pyramid graph transformer,” Pattern Recognition, vol. 137, pp. 109–117, May 2023.
[55] S.-L. Chang, L.-S. Chen, Y.-C. Chung, and S.-W. Chen, “Automatic license plate recognition,” IEEE Transactions on Intelligent Transportation Systems, vol. 5, no. 1, pp. 42–53, March 2004.
[56] T.-A. Pham, “Effective deep neural networks for license plate detection and recognition,” The Visual Computer, vol. 39, no. 3, pp. 927–941, March 2023.
[57] R. Li, S. Wang, P. Jiao, and S. Lin, “Traffic control optimization strategy based on license plate recognition data,” Journal of Traffic and Transportation Engineering, vol. 10, no. 1, pp. 45–57, February 2023.
[58] H. Padmasiri, J. Shashirangana, D. Meedeniya, O. Rana, and C. Perera, “Automated license plate recognition for resource-constrained environments,” Sensors, vol. 22, no. 4, pp. 143–154, February 2022.
[59] I. Dabral, M. Singh, and K. Kumar, “Cancer detection using convolutional neural network,” in Proc. International Conference on Deep Learning, Artificial Intelligence and Robotics, Jaipur, India, December 7–8, 2021, pp. 290–298.
[60] A. B. Nassif, M. A. Talib, Q. Nasir, Y. Afadar, and O. Elgendy, “Breast cancer detection using artificial intelligence techniques: A systematic literature review,” Artificial Intelligence in Medicine, vol. 127, pp. 102–126, May 2022.
[61] P. Xue, “Deep learning in image-based breast and cervical cancer detection: A systematic review and meta-analysis,” NPJ Digital Medicine, vol. 5, no. 1, pp. 19–29, February 2022.
[62] M. Nawaz, “Skin cancer detection from dermoscopic images using deep learning and fuzzy k-means clustering,” Microscopy Research and Technique, vol. 85, no. 1, pp. 339–351, January 2022.
[63] H. C. Reis, V. Turk, K. Khoshelham, and S. Kaya, “InSiNet: A deep convolutional approach to skin cancer detection and segmentation,” Medical, Biological Engineering and Computing, vol. 60, no. 3, pp. 643–662, March 2022.
[64] A. M. Burton, V. Bruce, and P. J. B. Hancock, “From pixels to people: A model of familiar face recognition,” Cognitive Science, vol. 23, no. 1, pp. 1–31, January 1999.
[65] A. Voulodimos, N. Doulamis, A. Doulamis, and E. Protopapadakis, “Deep learning for computer vision: A brief review,” Computational Intelligence and Neuroscience, vol. 2018, pp. 1–13, February 2018.
[66] A. Kumar, A. Kaur, and M. Kumar, “Face detection techniques: A review,” Artificial Intelligence Review, vol. 52, no. 2, pp. 927–948, August 2019.
[67] Z. Xie, J. Li, and H. Shi, “A face recognition method based on CNN,” Journal of Physics: Conference Series, vol. 1395, no. 1, pp. 12–26, November 2019.
[68] D. Wang, H. Yu, D. Wang, and G. Li, “Face recognition system based on CNN,” in Proc. International Conference on Computer Information and Big Data Applications, Guiyang, China, April 17–19, 2020, pp. 470–473.
[69] T. Bezdan and N. Bačanin Džakula, “Convolutional neural network layers and architectures,” in Proc. International Scientific Conference, Vladivostok, Russia, October 1–4, 2019, pp. 445–451.
[70] H. Ide and T. Kurita, “Improvement of learning for CNN with ReLU activation by sparse regularization,” in Proc. International Joint Conference on Neural Networks, Anchorage, USA, May 14–19, 2017, pp. 2684–2691.
[71] J. Gu, “Recent advances in convolutional neural networks,” Pattern Recognition, vol. 77, pp. 354–377, December 2015.
[72] A. F. Agarap, “Deep learning using rectified linear units (ReLU),” arXiv:1803.08375, pp. 1–7, March 2018.
[73] D. M. -White, B. Sattelberg, N. Blanchard, and R. Beveridge, “Exploring the interchangeability of CNN embedding spaces,” arXiv:2010.02323, pp. 1–11, October 2020.
[74] A. Parnami and M. Lee, “Learning from few examples: A summary of approaches to few-shot learning,” arXiv:2203.04291, pp. 1–32, March 2022.
[75] I. Drori, “A neural network solves, explains, and generates university math problems by program synthesis and few-shot learning at human level,” arXiv:2112.15594, pp. 1–181, May 2022.
[76] S. Jadon, “An overview of deep learning architectures in few-shot learning domain,” arXiv:2008.06365, pp. 1–11, August 2020.
[77] X. Li, Z. Sun, J.-H. Xue, and Z. Ma, “A concise review of recent few-shot meta-learning methods,” Neurocomputing, vol. 456, pp. 463–468, October 2021.
[78] M. Hou and I. Sato, “A closer look at prototype classifier for few-shot image classification,” arXiv:2110.05076, pp. 1–21, October 2021.
[79] S. Laenen and L. Bertinetto, “On episodes, prototypical networks, and few-shot learning,” arXiv:2012.09831, pp. 1–18, December 2020.
[80] G. S. Dhillon, P. Chaudhari, A. Ravichandran, and S. Soatto, “A baseline for few-shot image classification,” arXiv:1909.02729, pp. 1–20, September 2019.
[81] W. Li, “LibFewShot: A comprehensive library for few-shot learning,” arXiv:2109.04898, pp. 1–17, September 2021.
[82] D. Argüeso, “Few-shot learning approach for plant disease classification using images taken in the field,” Computers and Electronics in Agriculture, vol. 175, pp. 105–122, August 2020.
[83] N. Hilliard, L. Phillips, S. Howland, A. Yankov, C. D. Corley, and N. O. Hodas, “Few-shot learning with metric-agnostic conditional embeddings,” arXiv:1802.04376, pp. 1–8, February 2018.
[84] F. Lu, E. Raff, and F. Ferraro, “Neural bregman divergences for distance learning,” arXiv:2206.04763, pp. 1–16, June 2022.
[85] P. Sitikhu, K. Pahi, P. Thapa, and S. Shakya, “A comparison of semantic similarity methods for maximum human interpretability,” in Proc. Artificial Intelligence for Transforming Business and Society, Kathmandu, Nepal, November 5, 2019, pp. 1–4.
[86] S. Xia, Z. Xiong, Y. Luo, WeiXu, and G. Zhang, “Effectiveness of the euclidean distance in high dimensional spaces,” Optik, vol. 126, no. 24, pp. 5614–5619, December 2015.
[87] W. R. W. Mohd and L. Abdullah, “Similarity measures of pythagorean fuzzy sets based on combination of cosine similarity measure and euclidean distance measure,” AIP Conference Proceedings, vol. 1974, no. 1, pp. 3–17, June 2018.
[88] S. Kapoor, S. Khanna, and R. Bhatia, “Facial gesture recognition using correlation and mahalanobis distance,” arXiv:1003.1819, pp. 1–6, March 2010.
[89] K. L. Elmore and M. B. Richman, “Euclidean distance as a similarity metric for principal component analysis,” Monthly Weather Review, vol. 129, no. 3, pp. 540–549, March 2001.
[90] R. De Maesschalck, D. Jouan-Rimbaud, and D. L. Massart, “The mahalanobis distance,” Chemometrics and Intelligent Laboratory Systems, vol. 50, no. 1, pp. 1–18, January 2000.
[91] P. Bateni, R. Goyal, V. Masrani, F. Wood, and L. Sigal, “Improved few-shot visual classification,” arXiv:1912.03432, pp. 1–14, December 2019.
[92] O. Ledoit and M. Wolf, “The power of (non-)linear shrinking: A review and guide to covariance matrix estimation,” Journal of Financial Econometrics, vol. 20, no. 1, pp. 187–218, January 2022.
[93] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” arXiv:1512.03385, pp. 1–12, December 2015.
[94] S.-H. Hsiao and J.-S. R. Jang, “Improving ResNet-based feature extractor for face recognition via re-ranking and approximate nearest neighbor,” in Proc. IEEE International Conference on Advanced Video and Signal Based Surveillance, Madrid, Spain, November 29–December 2, 2019, pp. 1–8.
[95] S. Targ, D. Almeida, and K. Lyman, “Resnet in Resnet: Generalizing residual architectures,” arXiv:1603.08029, pp. 1–7, March 2016.
[96] T. Akiba, S. Suzuki, and K. Fukuda, “Extremely large minibatch SGD: Training ResNet-50 on ImageNet in 15 minutes,” arXiv:1711.04325, pp. 1–4, November 2017.
[97] H. Mikami, H. Suganuma, P. U-chupala, Y. Tanaka, and Y. Kageyama, “Massively distributed SGD: ImageNet/ResNet-50 training in a flash,” arXiv:1811.05233, pp. 1–7, November 2018.
[98] Y. You, Z. Zhang, C.-J. Hsieh, J. Demmel, and K. Keutzer, “ImageNet training in minutes,” in Proc. International Conference on Parallel Processing, New York, USA, August 13–16, 2018, pp. 1–10.
[99] M. Guillaumin and V. Ferrari, “Large-scale knowledge transfer for object localization in imagenet,” in Proc. IEEE Conference on Computer Vision and Pattern Recognition, Providence, USA, June 16–21, 2012, pp. 3202–3209.
[100] A. Veit, T. Matera, L. Neumann, J. Matas, and S. Belongie, “COCO-text: dataset and benchmark for text detection and recognition in natural images,” arXiv:1601.07140, pp. 1–8, January 2016.
[101] S.-H. Hsiao and J.-S. R. Jang, “Improving ResNet-based feature extractor for face recognition via re-ranking and approximate nearest neighbor,” in Proc. IEEE International Conference on Advanced Video and Signal Based Surveillance, Taipei, Taiwan, September 18–21, 2019, pp. 1–8.
[102] D. Zhou, Q. Hou, Y. Chen, J. Feng, and S. Yan, “Rethinking bottleneck structure for efficient mobile network design,” arXiv:2007.02269, pp. 1–24, July 2020.
[103] P. Foret, A. Kleiner, H. Mobahi, and B. Neyshabur, “Sharpness-aware minimization for efficiently improving generalization,” arXiv:2010.01412, pp. 1–20, October 2020.
[104] S. Xin and H. Liu, “Few-shot classification based on CBAM and prototype network,” in Proc. International Conference on Data-driven Optimization of Complex Systems, Chengdu, China, October 28–30, 2022, pp. 1–6.
[105] S. Sengupta, J.-C. Chen, C. Castillo, V. M. Patel, R. Chellappa, and D. W. Jacobs, “Frontal to profile face verification in the wild,” in Proc. IEEE Winter Conference on Applications of Computer Vision, New York, USA, March 7–10, 2016, pp. 1–9.
[106] Z. Liu, P. Luo, X. Wang, and X. Tang, “Deep learning face attributes in the wild,” arXiv:1411.7766, pp. 1–11, November 2014.
[107] M. Du, F. Yang, N. Zou, and X. Hu, “Fairness in deep learning: A computational perspective,” IEEE Intelligent Systems, vol. 36, no. 4, pp. 25–34, July 2021.
[108] D. Leslie, “Understanding bias in facial recognition technologies,” arXiv:2010.07023, pp. 1–49, October 2020.
[109] K. Zhang, Z. Zhang, Z. Li, and Y. Qiao, “Joint face detection and alignment using multitask cascaded convolutional networks,” IEEE Signal Processing Letters, vol. 23, no. 10, pp. 1499–1503, October 2016.
[110] A. Paszke, “PyTorch: An imperative style, high-performance deep learning library,” arXiv:1912.01703, pp. 1–12, December 2019.
[111] S. M. R. Arnold, P. Mahajan, D. Datta, I. Bunner, and K. S. Zarkias, “learn2learn: A library for meta-learning research,” arXiv:2008.12284, pp. 1–10, August 2020.
[112] M. Abadi, “TensorFlow: A system for large-scale machine learning,” arXiv:1605.08695, pp. 1–18, May 2016.
[113] A. Buslaev, V. I. Iglovikov, E. Khvedchenya, A. Parinov, M. Druzhinin, and A. A. Kalinin, “Albumentations: Fast and flexible image augmentations,” Information, vol. 11, no. 2, pp. 125–135, February 2020.