Age-Invariant Face Recognition to Aid Visually Impaired People

• Gonçalo Ferreira Neto UFPI
• Rodrigo Veras UFPI
• Kelson Aires UFPI
• Laurindo Britto Neto UFPI

DOI: https://doi.org/10.5753/eniac.2021.18244

Abstract

This paper proposes an approach to assist people with visual impairments in recognizing people regardless of age. The goal is to develop a system that uses a face recognition approach, focusing on age invariance, that returns good results compared to the results obtained in the literature review. The approach studied uses deep Convolutional Neural Networks CCNs, pretrained by the VGGFace2 dataset, to extract feature descriptors from face images and classify with the Linear SVM classification algorithm. As can be seen throughout the paper, the approach returned 89.9% accuracy using the FG-NET dataset with 1002 images. And using the CACD dataset, which contains 163,446 images divided into four different subsets, three sets for training and one for testing, the approach returned 85.2%, 82.4%, and 88.2% accuracy for each model trained with a different training set.

References

Ali, A. S. O., Sagayan, V., Saeed, A. M., Ameen, H., and Aziz, A. (2015). Age-invariant face recognition system using combined shape and texture features. IET Biometrics, 4(2):98–115.

Babu, K. and Kunju, N. (2017). Cross age face recognition using deep learning model based on center symmetric lps. ICECDS, pages 1897–1900.

Barkan, O., Weill, J., Wolf, L., and Aronowitz, H. (2013). Fast high dimensional vector multiplication face recognition. In IEEE ICCV, pages 1960–1967.

Britto Neto, L., Grijalva, F., Maike, V. R. M. L., Martini, L. C., Florencio, D., Baranauskas, M. C. C., Rocha, A., and Goldenstein, S. (2017). A kinect-based wearable face recognition system to aid visually impaired users. IEEE THMS, 47(1):52–64.

Britto Neto, L., Maike, V. R. M. L., Koch, F. L., Baranauskas, M. C. C., Rocha, A., and Goldenstein, S. (2015). A wearable face recognition system built into a smartwatch and the visually impaired user. In ICEIS, pages 5–12. INSTICC, SciTePress.

Britto Neto, L. S. (2016). Sistemas Wearable baseados em Métodos de Visão Computacional para Auxiliar Pessoas com Deficiência Visual. PhD thesis, UNICAMP, Campinas.

Cao, Q., Shen, L., Xie, W., Parkhi, O. M., and Zisserman, A. (2018). Vggface2: A dataset for recognising faces across pose and age. In IEEE FG, pages 67–74.

Chen, B., Chen, C., and Hsu, W. H. (2015). Face recognition and retrieval using cross-age reference coding with cross-age celebrity dataset. IEEE Trans. Multi., 17(6):804–815.

Chen, B.-C., Chen, C.-S., and Hsu, W. H. (2014). Cross-age reference coding for ageinvariant face recognition and retrieval. ECCV, pages 768–783.

Essa, E. (2016). Sparse random encoder for age invariant face recognition. 2016 11th International Conference on Computer Engineering Systems (ICCES), pages 167–171.

Gong, D., Li, Z., Lin, D., Liu, J., and Tang, X. (2013). Hidden factor analysis for age invariant face recognition. 2013 IEEE ICCV, pages 2872–2879.

Gong, D., Li, Z., Tao, D., Liu, J., and Li, X. (2015). A maximum entropy feature descriptor for age invariant face recognition. IEEE CVPR, pages 5289–5297.

Hou, X., Ding, S., and Ma, L. (2016). Robust feature encoding for age-invariant face recognition. IEEE ICME, pages 1–6.

Joachims, T. (2006). Training linear svms in linear time. In ACM SIGKDD, KDD ’06, pages 217–226, New York, NY, USA. ACM.

Kohavi, R. (1995). A study of cross-validation and bootstrap for accuracy estimation and model selection. IJCAI, pages 1137–1143.

Li, Y., Wang, G., Lin, L., and Chang, H. (2015). A deep joint learning approach for age invariant face verification. In Zha, H., Chen, X., Wang, L., and Miao, Q., editors, Computer Vision, pages 296–305, Berlin, Heidelberg. Springer Berlin Heidelberg.

Li, Z., Park, U., and Jain, A. K. (2011). A discriminative model for age invariant face recognition. IEEE TIFS, 6(3):1028–1037.

Panis, G., Lanitis, A., Tsapatsoulis, N., and Cootes, T. F. (2016). Overview of research on facial ageing using the fg-net ageing database. IET Biometrics, 5(2):37–46.

Park, U., Tong, Y., and Jain, A. K. (2010). Age-invariant face recognition. IEEE TPAMI, 32(5):947–954.

Pirlo, G. and Impedovo, D. (2013). Cosine similarity for analysis and verification of static signatures. IET Biometrics, 2(4):151–158.

Ricanek, K. and Tesafaye, T. (2006). Morph: a longitudinal image database of normal adult age-progression. FGR, pages 341–345.

Somada, Y., Ohyama, W., and Wakabayashi, T. (2017). Segmented face image verification for age-invariant face recognition. ICIEV-ISCMHT, pages 1–4.

Sungatullina, D., Lu, J., Wang, G., and Moulin, P. (2013). Multiview discriminative learning for age-invariant face recognition. IEEE FG, pages 1–6.

Wen, Y., Li, Z., and Qiao, Y. (2016). Latent factor guided convolutional neural networks for age-invariant face recognition. IEEE CVPR, pages 4893–4901.

Yang, H., Huang, D., and Wang, Y. (2014). Age invariant face recognition based on texture embedded discriminative graph model. IEEE IJCB, pages 1–8.

Yi, D., Lei, Z., Liao, S., and Li, S. Z. (2014). Learning face representation from scratch. CoRR, abs/1411.7923.

Yu, J. and Jing, L. (2018). A joint multi-task cnn for cross-age face recognition. In 2018 25th IEEE International Conference on Image Processing (ICIP), pages 2411–2415.

Zhao, J., Cheng, Y., Cheng, Y., Yang, Y., Lan, H., Zhao, F., Xiong, L., Xu, Y., Li, J., Pranata, S., Shen, S., Xing, J., Liu, H., Yan, S., and Feng, J. (2018). Look across elapse: Disentangled representation learning and photorealistic cross-age face synthesis for age-invariant face recognition. CoRR, abs/1809.00338.

Zheng, T., Deng, W., and Hu, J. (2017). Age estimation guided convolutional neural network for age-invariant face recognition. IEEE CVPRW, pages 503–511.