Incept-N: A Convolutional Neural Network based Classification Approach for Predicting Nationality from Facial Features
1. Paper Id- 254, RTIP2R-2018, Solapur University, Maharashtra state, INDIA
Masum Shah Junayed, Afsana Ahsan Jeny, Nafis Neehal, Eshtiak Ahmed,
Syed Akhter Hossain.
Department of Computer Science and Engineering.
Daffodil International University, Dhaka, Bangladesh.
Incept-N: A Convolutional Neural
Network based Classification
Approach for Predicting Nationality
from Facial Features
3. Overview:Why we proposed Incept-N?
Nationality of a human being is a well-known
identifying characteristic used for every major
authentication purpose in every country. With a
goal to successfully applying computer vision
techniques to predict a humans nationality based
on his facial features, we have proposed this
novel method.
3
4. 4
Literature Review
Till now, it has been found that there are plethora of works
done on nationality identification. A few works done on various
facial recognition using ImageNet [2], machine learning.
• Effective Computer Model for Recognizing Nationality from
Frontal Image [4]. They used SVM, AAM, ASM and the
accuracy was 86.4%. Their experiment was worked manually
and images must be the frontal face image that has smooth
lighting and does not have any rotation angle.
• Facial Expression Recognition based on the Inception-v3 [8]
model of TensorFlow [1] platform. Their accuracy was 97%
but it wasnt based on dynamic sequences [7].
• Ethnicity Identification from Face Images[11].
6. Collect data from real
life, Open Source web.
We have collected 600
images with different
angle of Five different
countries from facial
feature.
Resize Dataset.
Incept-N : Dataset
6
Figure 1: The sample of our dataset.
7. 7
Incept-N : Dataset Preparation
Used 5 Augmented methods
Rotate right +30 degree,
Rotate left -30 degree,
Flip horizontally,
Shading
Translation
12. Result Discussion
Figure 1: The variation of
accuracy on Incept-N Dataset.
Figure 2: The variation of cross
entropy on Incept-N Dataset.
12
13. Future Work
Complete and enrich the dataset.
Developing real life Application .
Building own CNN from scratch to improve accuracy.
Evaluate the best Neural Network by applying several models
like VGG-16, AlexNet, ResNet etc.
13
14. References
1. Martin Abadi, Ashish Agarwal, et aI, TensorFlow: Large-Scale Machine Learning on Heterogeneous
Distributed Systems. CoRR abs/1603.04467 , 2016.
2. Alex Krizhevsky Ilya Sutskever Geoffrey E. Hinton, “ImageNet Classification with Deep Convolutional
Neural Networks”, December 03 - 06, 2012.
3. Alwyn Mathew*a, Jimson Mathew, Mahesh Govind, Asif Mooppanb, “An Improved Transfer learning
Approach for Intrusion Detection”, 7th International Conference on Advances in Computing &
Communications, ICACC-2017, 22-24 August 2017, Cochin, India.
4. Bat-Erdene.B and Ganbat.Ts, “Effective Computer Model For Recognizing Nationality From Frontal Image”.
5. Xiaoling Xia and Cui Xu, “Inception-v3 for Flower Classification”, 2017 2nd International Conference on
Image, Vision and Computing.
6. Brady Kieffer1, Morteza Babaie2 Shivam Kalra1, and H.R.Tizhoosh1, “Convolutional Neural Networks for
Histopathology Image Classification: Training vs. Using Pre-Trained Networks”, arXiv:1710.05726v1
[cs.CV] 11 Oct 2017.
7. Xiao-Ling Xia 1, Cui Xu*2, Bing Nan3, “Facial Expression Recognition Based on TensorFlow Platform”, ITM
Web of Conferences 12, 01005 (2017).
8. Christian Szegedy Vincent Vanhoucke Sergey Ioffe Jonathon Shlens Zbigniew Wojna, “Rethinking the
Inception Architecture for Computer Vision”, arXiv: 1512.00567v1 [cs.CV] 2 Dec 2015.
9. https://datascience.stackexchange.com/questions/15328/what-is-thedifferencebetween-inception-v2-
and-inception-v3.
10.https://codelabs.developers.google.com/codelabs/tensorflow-forpoets/#0
11.Xiaoguang Lu and Anil K. Jain, “Ethnicity Identification from Face Images”, Proceedings Volume 5404,
Biometric Technology for Human Identification; (2004).
12.Kanis Charntaweekhun and Somkiat Wangsiripitak, “Visual Programming using Flowchart”, 2006
International Symposium on Communications and Information Technologies.
13.https://datascience.stackexchange.com/questions/15989/microaverage-vs-macroaverage-performance-
in-a-multiclass-classification
14.B.P. Salmon, W. Kleynhans, C.P. Schwegmann and J.C. Olivier, “Proper comparison among methods using
a confusion matrix”, 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS).
14
Sports activities are an integral part of our day to day life. Recognizing and analysing different sports events and activities has become an emerging trend in computer vision arena.