DEEP FACE - On the Reconstruction of Face Images from Deep Face Templates

AUTHOR(S)

Amal Joseph, Binny S, Abhishek V A, Nithin Raj, Vimel Manoj

DOI: https://doi.org/10.46647/ijetms.2023.v07i04.083

ABSTRACT
The paper on “Reconstruction of Face Images from Deep Face Templates" presents a novel approach for face image reconstruction using deep learning techniques. The proposed method utilizes a pre-trained deep face template, which is a convolutional neural network (CNN) trained on a large-scale face dataset, as a prior to guide the reconstruction process. Specifically, the method solves an optimization problem that balances the fidelity to the input image and the similarity to the deep face template. Its then evaluated with the method on two face image datasets, and demonstrate that their method outperforms several state-of-the-art methods in terms of reconstruction quality, especially for images with large occlusions or low resolutions. Moreover, they show that the deep face template can capture high-level face attributes, such as pose, identity, and expression, which can be used for various face-related tasks, such as face recognition, attribute manipulation, and generation. Overall, the paper presents a promising direction for face image reconstruction using deep learning techniques, and highlights the potential of deep face templates for capturing and utilizing high-level face attributes.

Page No: 606 - 611

References:

1. [1] A. Adler, “Sample images can be independently restored from face recognition templates,” in
   CCECE, 2003.
   [2] J. Galbally, C. McCool, J. Fierrez, S. Marcel, and J. Ortega-Garcia, “On the vulnerability of
   face verification systems to hill-climbing attacks,” Pattern Recognition, 2010.
   [3] Y. C. Feng, M.-H. Lim, and P. C. Yuen, “Masquerade attack on transform-based binary-template
   protection based on perceptron learning,” Pattern Recognition, 2014.
   [4] D. Wen, H. Han, and A. K. Jain, “Face spoof detection with image distortion analysis,” IEEE
   Transactions on Information Forensics and Security, 2015.
   [5] K. Patel, H. Han, and A. K. Jain, “Secure face unlock: Spoof detection on smartphones,” IEEE
   Transactions on Information Forensics and Security, 2016.
   [6] S. Liu, P. C. Yuen, S. Zhang, and G. Zhao, “3d mask face anti spoofing with remote
   photoplethysmography,” in ECCV, 2016.
   [7] R. Shao, X. Lan, and P. C. Yuen, “Deep convolutional dynamic texture learning with adaptive
   channel-discriminability for 3d mask face anti-spoofing,” in IJCB, 2017.
   [8] Y. Liu, A. Jourabloo, and X. Liu, “Learning deep models for faceanti-spoofing: Binary or
   auxiliary supervision,” in CVPR, 2018.
   [9] A. Mignon and F. Jurie, “Reconstructing faces from their signatures using rbf regression,” in
   BMVC, 2013.
   [10] “Face id security,” Apple Inc, 2017. [Online]. Available: ”https:// images.apple.com/business/docs/FaceID Security Guide.pdf”
   [11] P. Mohanty, S. Sarkar, and R. Kasturi, “From scores to face templates: a model-based
   approach,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007.
   [12] S. U. Hussain, T. Napol´eon, and F. Jurie, “Face recognition using local quantized patterns,”
   in BMVC, 2012.
   [13] A. Zhmoginov and M. Sandler, “Inverting face embeddings with convolutional neural
   networks,” arXiv:1606.04189, 2016.
   [14] F. Schroff, D. Kalenichenko, and J. Philbin, “Facenet: A unified embedding for face
   recognition and clustering,” in CVPR, 2015.
   [15] F. Cole, D. Belanger, D. Krishnan, A. Sarna, I. Mosseri, and W. T. Freeman, “Synthesizing
   normalized faces from facial identity features,” in CVPR, 2017




How to Cite This Article:
Amal Joseph, Binny S, Abhishek V A, Nithin Raj, Vimel Manoj . DEEP FACE - On the Reconstruction of Face Images from Deep Face Templates . ijetms;7(4):606-611. DOI: 10.46647/ijetms.2023.v07i04.083