Event Title

Improvement of Gender Recognition using the Cosfire Filter Framework (Simulations Platform of Shape-Preserving Regression – PCHIP)

Session

Computer Science and Communication Engineering

Description

Biometrics is evolving every day more and more in technical sense and consequently faces with further challenges that become sharper. One of these challenges of is gender recognition that finds very important and key applications. In this paper, we consider the gender recognition process implemented through the Cosfire filter applied through Viola-Jones algorithm and simulated through the Matlab platform. Objective of this paper is improving the execution of gender recognition. The database contains 237 images of 128 to 128 pixels, where 128 are males and 109 are females. For each of them, gender recognition is performed by applying current and improved Viola-Jones algorithm and execution time for each of them is measured. Consequently, it is noticed that the execution time in the case of modified algorithm is lower than the first version. The change consists in intervening in recursive filtering by duplicating it. Furthermore, data obtained from both algorithms in question are processed through the Shape-Preserving Regression - PCHIP regression by giving respective equations and the coefficients of the determination and the respective residual plots performed by Matlab simulation test-bench. Recommendations can be issued in context of further execution time reduction of Viola-Jones algorithm applied on gender recognition.

Keywords:

biometrics, execution, Viola-Jones, filtering

Session Chair

Bertan Karahoda

Session Co-Chair

Krenare Pireva

Proceedings Editor

Edmond Hajrizi

ISBN

978-9951-550-19-2

Location

Pristina, Kosovo

Start Date

26-10-2019 1:30 PM

End Date

26-10-2019 3:00 PM

DOI

10.33107/ubt-ic.2019.267

This document is currently not available here.

Share

COinS
 
Oct 26th, 1:30 PM Oct 26th, 3:00 PM

Improvement of Gender Recognition using the Cosfire Filter Framework (Simulations Platform of Shape-Preserving Regression – PCHIP)

Pristina, Kosovo

Biometrics is evolving every day more and more in technical sense and consequently faces with further challenges that become sharper. One of these challenges of is gender recognition that finds very important and key applications. In this paper, we consider the gender recognition process implemented through the Cosfire filter applied through Viola-Jones algorithm and simulated through the Matlab platform. Objective of this paper is improving the execution of gender recognition. The database contains 237 images of 128 to 128 pixels, where 128 are males and 109 are females. For each of them, gender recognition is performed by applying current and improved Viola-Jones algorithm and execution time for each of them is measured. Consequently, it is noticed that the execution time in the case of modified algorithm is lower than the first version. The change consists in intervening in recursive filtering by duplicating it. Furthermore, data obtained from both algorithms in question are processed through the Shape-Preserving Regression - PCHIP regression by giving respective equations and the coefficients of the determination and the respective residual plots performed by Matlab simulation test-bench. Recommendations can be issued in context of further execution time reduction of Viola-Jones algorithm applied on gender recognition.