An Accuracy Improvement in Facial Verification through Siamese Neural Networks
Keywords:
Siamese Networks, CNN, Loss Functions, Hyper-Parameters, Image RecognitionAbstract
Due to the ever-growing number of digital photos on the Internet, the task of automatically analysing images to extract semantic content has become quite important in recent years. In order to successfully organize the photographs, image indexing and image retrieval strategies need top-notch, effective image analysis and pattern recognition algorithms that can extract relevant semantic data. Face recognition plays a critical role in a variety of real-world applications, including AI-driven access control, interactions, immigration, education, and retail. All these industries have successfully used machine learning. However, the applications can be computationally expensive and may be challenging to perform when limited data is available. This is where the grey area exists. In this study, we provide a Siamese neural network-based technique for facial verification. To categorize the similarity between two input faces, a neural network is used. The Siamese neural network was successfully used in earlier experiments to recognize characters. This study will attempt to expand on the research into facial recognition software. The network can predict fresh data as well as new classes from unknown distributions once it has been tuned. In this study, convolutional neural networks were employed to exceed predictions made using previously developed methods.
References
Zagoruyko S, Komodakis N. Learning to compare image patches via convolutional neural networks. In Proceedings of the IEEE conference on computer vision and pattern recognition. 2015; 4353–4361.
Xu XF, Zhang L, Duan CD, Lu Y. Research on inception module incorporated siamese convolutional neural networks to realize face recognition. IEEE Access. 2019; 8: 12168–12178.
Shen Y, Wang H, Dai Y. Deep siamese network-based classifier and its application. Computer Engineering Applications. 2018; 54(10): 19–25.
Chan TH, Jia K, Gao S, Lu J, Zeng Z, Ma Y. PCANet: A simple deep learning baseline for image classification. IEEE Trans Image Process. 2015 Dec; 24(12): 5017–5032.
Bromley J, Guyon I, LeCun Y, Sa¨ckinger E, Shah R. Signature verification using a “siamese” time delay neural network. In: Advances in neural information processing systems. 1993; 6.
Koch G, Zemel R, Salakhutdinov R. Siamese neural networks for one-shot image recognition. In ICML deep learning workshop. 2015 Jul; 2(1).
Taigman Y, Ranzato MA, et al. DeepFace: Closing the Gap to Human-Level Performance in Face Verification. In 2014 IEEE Conference on Computer Vision and Pattern Recognition, CVPR Columbus, OH, USA. 2014; 1701–1708.
Chong W, Blei D, Li FF. Simultaneous image classification and annotation. In 2009 IEEE Conference on computer vision and pattern recognition. 2009 Jun; 1903–1910.
Liu X, Zhou Y, Zhao J, Yao R, Liu B, Zheng Y. Siamese convolutional neural networks for remote sensing scene classification. IEEE Geosci Remote Sens Lett. 2019 Aug; 16(8): 1200–1204.
Vinyals O, Blundell C, Lillicrap T, Wierstra D. Matching networks for one shot learning. Advances in Neural Information Processing Systems. 2016; 29.
Bukovcikova Z, Sopiak D, Oravec M, Pavlovicova J. Face verification using convolutional neural networks with Siamese architecture. In Proc Int Symp (ELMAR). 2017 Sep; 205–208.
Schroff F, Kalenichenko D, Philbin J. Facenet: A unified embedding for face recognition and clustering. In Proceedings of the IEEE conference on computer vision and pattern recognition. 2015; 815–823.
Dhillon GS, Chaudhari P, Ravichandran A, Soatto S. A baseline for few-shot image classification. arXiv preprint arXiv:1909.02729. 2019.
Figueroa-Mata G, Mata-Montero E. Using a Convolutional Siamese Network for Image-Based Plant Species Identification with Small Datasets. Biomimetics (Basel). 2020 Mar 1; 5(1): 8. doi: 10.3390/biomimetics5010008. PMID: 32121572; PMCID: PMC7148474.
Schroff F, Kalenichenko D, Philbin J. Facenet: A unified embedding for face recognition and clustering. In Proceedings of the IEEE conference on computer vision and pattern recognition. 2015; 815–823.
Atoum Y, Liu Y, Jourabloo A, Liu X. Face anti-spoofing using patch and depth-based CNNs. In Proc IEEE Int Joint Conf Biometrics (IJCB). 2017 Oct; 319–328.
Chen WY, Liu YC, Kira Z, Wang YCF, Huang JB. A closer look at few-shot classification. arXiv preprint arXiv:1904.04232. 2019.
Chen Y, Wang X, Liu Z, Xu H, Darrell T. A new meta-baseline for few-shot learning. arXiv:2003.04390v2 [cs.CV]. 2020 Apr.
Lungu IA, Hu Y, Liu SC. Multi- resolution siamese networks for one-shot learning. In 2020 2nd IEEE International Conference on Artificial Intelligence Circuits and Systems (AICAS). 2020 Aug; 183–187.
Vargas C, Zhang Q, Izquierdo E. One shot logo recognition based on siamese neural networks. In Proceedings of the 2020 International Conference on Multimedia Retrieval. 2020 Jun; 321–325.
Toscani M, Gil R, Guarnera DY, Guarnera G, Kalouaz A, Gegenfurtner KR. Assessment of OLED head mounted display for vision research with virtual reality. In 2019 IEEE 15th International Conference on Signal- Image Technology Internet-Based Systems (SITIS). 2019 Nov; 738–745.
Wu K, Wang P, Wang Z. Few-shot malicious traffic classification based on Siamese Neural Network. In 2021 IEEE 23rd Int Conf on High Performance Computing Communications; 7th Int Conf on Data Science Systems; 19th Int Conf on Smart City; 7th Int Conf on Dependability in Sensor, Cloud Big Data Systems Application (HPCC/DSS/SmartCity/DependSys). 2021 Dec; 1670–1675. IEEE.
