CLASSIFICATION OF GENDER USING FUSION AND LBP FEATURES

Project Abstract / Summary : Abstract— In this project, we report our extension of the use of feature selection based on mutual information and feature fusion to improve gender classification of face images. We compare the results of fusing three groups of features, three spatial scales, and four different mutual information measures to select features. We also showed improved results by fusion of LBP features with different radii and spatial scales, and the selection of features using mutual information. As measures of mutual information we use minimum redundancy and maximal relevance (mRMR), normalized mutual information feature selection (NMIFS), conditional mutual information feature selection (CMIFS), and conditional mutual information maximization (CMIM). We also show a significant reduction in processing time because of the feature selection, which makes real-time applications of gender classification feasible.


Index Terms—Feature fusion, feature selection, gender classification, local binary patterns, mutual inform
I. Introduction

feature selection plays an important role in improving accuracy, efficiency and scalability of the object identification process. Since relevant features are often unknown a priori in the real world, irrelevant and redundant features may be introduced to represent the domain. However, using more features implies increasing computational cost in the feature extraction process, slowing down the classification process and also increasing the time needed for training and validation, which may lead to classification over-fitting.




Why did you choose to work on this project topic : The feature selection methods used in this paper act as filters eliminating most of the features with low relevance or high redundancy and provide an efficient approach in terms of the computational time required for gender classification. These methods are considered effective for feature selection, especially when a large number of features are processed. In our model, the process of training the SVM classifier was achieved more efficiently and effectively eliminating a significant number of features with low relevance and high redundancy.

Project Category : Electrical / Electronics / Communication
------------------------------------------------------
Institute/College Name: PALLAVAN COLLEGE OF ENGINEERING
City: KANCHIPURAM
State: TAMILNADU
Participating Team From: Final Year

Replies

You are reading an archived discussion.

Related Posts

The Indian mobile manufacturer Intex has unveiled another Aqua series smartphone named Aqua Power HD. Back in December last year, the company had launched Aqua Power with a huge 4,000...
Hello, I am trying to install and start learning openFOAM CFD myself, on a windows system. The installation for this predominantly (Linux based) open source CFD environment has been given...
With so many unconventional launches in CES 2015, there were quite a many products which didn’t receive the attention they should have. Canon PowerShot SX530 is one of such products....
Acer India debuted their new creation Acer One 10 (S1001) as an Amazon India exclusive recently. This Windows 8.1 device works as a standalone tablet and when paired with its...
i am doing project in image segmentation using isoperimetric partitioning. i dont know how to do this in matlab. kindly let me know if anyone know about this