Page - 151 - in Proceedings of the OAGM&ARW Joint Workshop - Vision, Automation and Robotics

(a)FCNNtrainedwith synthetic data (b) FCNN trained with augmented data (c) FCNN trained on augmented + GAN data. Fig. 11. Comparison of the output of the same network trained only on synthetic, on augmented and on refined data. Fig. 12. 0% match of two impressions of the same finger taken from FVC 2002 [16] database with minutiae extracted using the FCNN algorithm. VI. CONCLUSION In this work the possibility of reformulating the fingerprint minutiae extraction problem as a binary segmentation task is shown. Deep learning is used to address this problem. Even with synthetic data as a substitute to annotated real data, the algorithm is able to detect reasonable minutiae with better results than MINDTCT on the FVC2000 dataset without fine tuning of any parameters. Additionally, the performance gain of using our refinement approach was clearly illustrated and advances in training GANs are likely to bring better performance for this minutiae extraction algorithm. A first step is made by using the Hessian instead of the image itself for regularization. However, this performance gain illustrates the dependence on good training data. Currently, the angle of the minutiae points are calculated using an orientation field. In a future network, we want to learn the orientation of the minutiae by using the orientation field of the ground truth ridge pattern. We believe that better than state-of-the-art performance can be reached using deep learning given sufficiently diverse training data. ACKNOWLEDGMENT We thank Peter Wild and Thomas Pock for their helpful insights. REFERENCES [1] “UareU Database,” http://www.neurotechnology.com/download.html, 2007, [Online; accessed 01-March-2017]. [2] A. H. Ansari, “Generation and storage of large synthetic fingerprint database,” Ph.D. dissertation, Indian Institute of Science Bangalore, 2011. [3] M. Arjovsky and L. Bottou, “Towards principled methods for training generative adversarial networks,” in NIPS 2016 Workshop on Adver- sarial Training. In review for ICLR, vol. 2016, 2017. [4] K. Cao, E. Liu, and A. K. Jain, “Segmentation and enhancement of latent fingerprints: A coarse to fine ridgestructure dictionary,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 36, no. 9, pp. 1847–1859, 2014. [5] R. Cappelli, D. Maio, and D. Maltoni, “Sfinge: an approach to syn- thetic fingerprint generation,” in International Workshop on Biometric Technologies (BT2004), 2004, pp. 147–154. [6] F. Chollet, “Keras,” https://github.com/fchollet/keras, 2015. [7] M.Drozdzal,E.Vorontsov,G.Chartrand,S.Kadoury, andC.Pal, “The importance of skip connections in biomedical image segmentation,” in International Workshop on Large-Scale Annotation of Biomedical Data and Expert Label Synthesis. Springer, 2016, pp. 179–187. [8] I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y. Bengio, “Generative adversarial nets,” in Advances in neural information processing systems, 2014, pp. 2672– 2680. [9] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” arXiv preprint arXiv:1512.03385, 2015. [10] L. Hong, Y. Wan, and A. Jain, “Fingerprint image enhancement: algorithm and performance evaluation,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 20, no. 8, pp. 777–789, 1998. [11] S. Ioffe and C. Szegedy, “Batch normalization: Accelerating deep network training by reducing internal covariate shift,” arXiv preprint arXiv:1502.03167, 2015. [12] S. Je´gou, M. Drozdzal, D. Vazquez, A. Romero, and Y. Bengio, “The one hundred layers tiramisu: Fully convolutional densenets for semantic segmentation,” arXiv preprint arXiv:1611.09326, 2016. [13] D. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980, 2014. [14] J. Long, E. Shelhamer, and T. Darrell, “Fully convolutional networks for semantic segmentation,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015, pp. 3431–3440. [15] D. Maio, D. Maltoni, R. Cappelli, J. L. Wayman, and A. K. Jain, “Fvc2000:Fingerprintverificationcompetition,” IEEETransactionson PatternAnalysisandMachine Intelligence, vol.24,no.3,pp.402–412, 2002. [16] ——, “Fvc2002: Second fingerprint verification competition,” in 16th international conference on Pattern Recognition. Proceedings., vol. 3. IEEE, 2002, pp. 811–814. [17] D. Maltoni, D. Maio, A. Jain, and S. Prabhakar, Handbook of fingerprint recognition. Springer Science & Business Media, 2009. [18] A. A. Moenssens, Fingerprint techniques. Chilton Book Company London, 1971. [19] O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in International Con- ference on Medical Image Computing and Computer-Assisted Inter- vention. Springer, 2015, pp. 234–241. [20] T. Salimans, I. Goodfellow, W. Zaremba, V. Cheung, A. Radford, and X. Chen, “Improved techniques for training gans,” in Advances in Neural Information Processing Systems, 2016, pp. 2226–2234. [21] A. Shrivastava, T. Pfister, O. Tuzel, J. Susskind, W. Wang, and R. Webb, “Learning from simulated and unsupervised images through adversarial training,” arXiv preprint arXiv:1612.07828, 2016. [22] C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, and Z. Wojna, “Rethink- ing the inception architecture for computer vision,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016, pp. 2818–2826. [23] Y. Tang, F. Gao, and J. Feng, “Latent fingerprint minutia extraction using fully convolutional network,” arXiv preprint arXiv:1609.09850, 2016. [24] S. VeriFinger, “Neuro technology (2010),” 2010. [25] C. I. Watson, M. D. Garris, E. Tabassi, C. L. Wilson, R. M. Mccabe, S. Janet, and K. Ko, “User’s guide to nist biometric image software (nbis),” 2007. [26] C. I. Watson and C. Wilson, “Nist special database 4,” Fingerprint Database, National Institute of Standards and Technology, vol. 17, p. 77, 1992. [27] S. Zagoruyko and N. Komodakis, “Wide residual networks,” CoRR, vol. abs/1605.07146, 2016. [Online]. Available: http://arxiv.org/abs/1605.07146 151