The week's pick
Random Articles
Reseach Article
Facial Recognition using Unsupervised Mode in Artificial Neural Network by Self-Organizing Map using MATLAB
| Communications on Applied Electronics |
| Foundation of Computer Science (FCS), NY, USA |
| Volume 5 - Number 5 |
| Year of Publication: 2016 |
| Authors: Harsha Kosta |
10.5120/cae2016652319
|
Harsha Kosta . Facial Recognition using Unsupervised Mode in Artificial Neural Network by Self-Organizing Map using MATLAB. Communications on Applied Electronics. 5, 5 ( Jul 2016), 29-32. DOI=10.5120/cae2016652319
Abstract
The main aim of this paper is to label a Self-Organizing Map (SOM) to measure image similarity. To achieve this, we input Facial images linked to the regions of interest into the neural network. Every single neural unit is tuned to a particular Facial image prototype at the end of the learning step. Probabilistic decision rule then performs facial recognition. This method gives accurate results for face identification taking into consideration various factors such as illumination variation and facial poses and expressions. The current research reveals a promising Self-Organizing Map (SOM) for face recognition. The Self-Organizing Map (SOM) technique is trained on images from one database. The novelty of this research comes from the integration of Images from input database, Training and Mapping. Face Recognition using unsupervised mode in neural network by Self-Organizing Map. Out of the all the architectures and algorithms suggested for artificial neural network, the SOM has an advantageous property of effectively creating a spatially organized `internal representation’ of various features of input signals and their abstractions. After the weight vectors have been finely tuned, the SOM technique has been promisingly accurate in various pattern recognition tasks involving very noisy signal. One develops realistic cortical structures when given approximations of visual environment as input, and is effective way to model the development of face recognition capability. In this study Kohonen self-organizing map (SOM) based retrieval system has been used to develop and illustrate a recognition system for human faces. It is due to topological ordering that Self-organizing map possess a better and precise feature recognizing and extracting property. Factoring in the Facial Analysis solutions for over 400 images used from the AT&T database we can say safely say that the facial recognition rate using one of the neural network algorithm SOM is 92.40% for 40 persons.
References
- G. B. Huang, M. Ramesh, T. Berg and E. Learned-Miller, "Labeled faces in the wild: A database for studying face recognition in unconstrained environments", 2007.
- R. Beveridge and H. Zhang, "Report on the FG 2015 video person recognition evaluation", Proc. IEEE Int. Conf. Autom. Face Gesture Recog., pp. 1-8, 2015.
- C. Ding and D. Tao, "A comprehensive survey on pose-invariant face recognition,"arXiv preprint arXiv:1502.04383, 2015.
- Santaji Ghorpade, Jayshree G., Shamla Mantri. (Dec 2010): Pattern Recognition using Neural Networks. IJCSIT, ISSN- 0975-9646, Vol 2, No. 6, pp92-98.
- Santaji Ghorpade, Jayshree G., Shamla Mantri, Dhanaji G. (Dec 2010): Neural Networks for Face Recognition using SOM.IJCST, ISSN-0976-8491, Vol 1, issue 2.
- Teuvo Kohonen.( 1990), Self-Organizing Map, Vol. 78, No.9.
- Dinesh Kumar Dimensionality. (May 2008), Reduction using SOM based on Technique for Face Recognition, Journal of Multimedia, Vol.3, No.1.
- F, Nazish Jamil, Fatima Jinnah Women University. Rewalpindi, Pakistan ace Recognition using Neural Networks.
- AT&T Laboratories Cambridge, The database of faces at http://www.cl.cam.ac.uk/research/dtg/attarchive/facesataglance. html.
Index Terms
Keywords