Deep Learning Features for Face Age Estimation: Better Than Human?

Deep convolutional neural networks have the ability to infer highly representative features from data. We decided to use this power for the purpose of estimating the human face age from a single colour image. We trained the Support Vector Machine regression model on raw feature vectors from the FaceNet deep neural network pretrained for face recognition. Our proposed method is a simple but effective FaceNet extension which does not need large scale data. In order to measure the accuracy of our approach, we proposed a test procedure on the FACES database for which we achieved the mean absolute error of 5.18 and the mean error of 0.09 years. Then, we conducted an experiment employing 78 students and showed that our method outperforms human for faces in the regular upright orientation. For vertically inverted faces, we reported an age underestimation trend in responses of students and our method.

[1]  Xin Liu,et al.  AgeNet: Deeply Learned Regressor and Classifier for Robust Apparent Age Estimation , 2015, 2015 IEEE International Conference on Computer Vision Workshop (ICCVW).

[2]  Natalie C. Ebner,et al.  FACES—A database of facial expressions in young, middle-aged, and older women and men: Development and validation , 2010, Behavior research methods.

[3]  Yuxiao Hu,et al.  MS-Celeb-1M: Challenge of Recognizing One Million Celebrities in the Real World , 2016, IMAWM.

[4]  M. Eimer.  The Face-Sensitive N170 Component of the Event-Related Brain Potential , 2011 .

[5]  Yu Qiao,et al.  A Discriminative Feature Learning Approach for Deep Face Recognition , 2016, ECCV.

[6]  Catherine A. Sugar,et al.  Inverted face processing in body dysmorphic disorder. , 2010, Journal of psychiatric research.

[7]  T. Sejnowski,et al.  Irresistible environment meets immovable neurons , 1997, Behavioral and Brain Sciences.

[8]  William G Hayward,et al.  Holistic processing underlies gender judgments of faces , 2010, Attention, perception & psychophysics.

[9]  Timothy F. Cootes,et al.  Overview of research on facial ageing using the FG-NET ageing database , 2016, IET Biom..

[10]  Thomas S. Huang,et al.  Face age estimation using patch-based hidden Markov model supervectors , 2008, 2008 19th International Conference on Pattern Recognition.

[11]  Yun Fu,et al.  Image-Based Human Age Estimation by Manifold Learning and Locally Adjusted Robust Regression , 2008, IEEE Transactions on Image Processing.

[12]  Catherine J Mondloch,et al.  Aging Faces and Aging Perceivers: Young and Older Adults are Less Sensitive to Deviations from Normality in Older Than in Young Adult Faces , 2013, Perception.

[13]  Xiaoming Liu,et al.  Demographic Estimation from Face Images: Human vs. Machine Performance , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[14]  Natalie C. Ebner,et al.  Let me guess how old you are: effects of age, gender, and facial expression on perceptions of age. , 2012, Psychology and aging.

[15]  Hazim Kemal Ekenel,et al.  How Transferable Are CNN-Based Features for Age and Gender Classification? , 2016, 2016 International Conference of the Biometrics Special Interest Group (BIOSIG).

[16]  N. Ambady,et al.  Investigating the Early Stages of Person Perception: The Asymmetry of Social Categorization by Sex vs. Age , 2014, PloS one.

[17]  Michael S. Bernstein,et al.  ImageNet Large Scale Visual Recognition Challenge , 2014, International Journal of Computer Vision.

[18]  James Philbin,et al.  FaceNet: A unified embedding for face recognition and clustering , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[19]  Luc Van Gool,et al.  DEX: Deep EXpectation of Apparent Age from a Single Image , 2015, 2015 IEEE International Conference on Computer Vision Workshop (ICCVW).

[20]  Jürgen Schmidhuber,et al.  Multi-column deep neural network for traffic sign classification , 2012, Neural Networks.

[21]  Yu Qiao,et al.  Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks , 2016, IEEE Signal Processing Letters.

[22]  Yun Fu,et al.  A study on automatic age estimation using a large database , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[23]  Adrian Schwaninger,et al.  Expert face processing: specialisation and constraints , 2003 .

[24]  S Brédart,et al.  An ‘Other-Race’ Effect in Age Estimation from Faces , 2001, Perception.

[25]  J. Bartlett,et al.  Inversion and Configuration of Faces , 1993, Cognitive Psychology.

[26]  Niels da Vitoria Lobo,et al.  Age classification from facial images , 1994, 1994 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[27]  Daniel Fitousi,et al.  Binding sex, age, and race in unfamiliar faces: The formation of “face files” , 2017 .

[28]  Mahir Faik Karaaba,et al.  Deep Convolutional Neural Networks and Support Vector Machines for Gender Recognition , 2015, 2015 IEEE Symposium Series on Computational Intelligence.

[29]  Ottmar V Lipp,et al.  Facial age cues and emotional expression interact asymmetrically: age cues moderate emotion categorisation , 2018, Cognition & emotion.

[30]  Yan Li,et al.  A Study on Apparent Age Estimation , 2015, 2015 IEEE International Conference on Computer Vision Workshop (ICCVW).

[31]  Marwan Mattar,et al.  Labeled Faces in the Wild: A Database forStudying Face Recognition in Unconstrained Environments , 2008 .

[32]  Albina Saikauskienė,et al.  Face recognition peculiarities of different age people in a Lithuanian sample , 2015 .

[33]  A. Young,et al.  Orientation-sensitivity to facial features explains the Thatcher illusion. , 2014, Journal of vision.

[34]  P. Thompson,et al.  Margaret Thatcher: A New Illusion , 1980, Perception.

[35]  Yuxiao Hu,et al.  MS-Celeb-1M: A Dataset and Benchmark for Large-Scale Face Recognition , 2016, ECCV.