Detecting Human and Classification of Gender using Facial Images MSIFT Features Based GSVM
G D K Kishore1, BabuReddy Mukamalla2

1G D K Kishore, Research Scholar, Computer Science, Krishna University, India.
2 BabuReddy Mukamalla, Assistant Professor, Computer Science, Krishna University, India.

Manuscript received on 5 August 2019. | Revised Manuscript received on 12 August 2019. | Manuscript published on 30 September 2019. | PP: 1466-1471 | Volume-8 Issue-3 September 2019 | Retrieval Number: B3782078219/19©BEIESP | DOI: 10.35940/ijrte.B3782.098319
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: Classification of gender using face recognition system is an essential concept for different types of applications in human-computer interaction and computer-aided related applications. It defines a wide range of features from human images to detect male, female and others using real-time data. There are different machine learning approaches were implemented to classify gender and also detects other images during the classification phase, which are not humans based on features extracted from human images datasets. All these existing techniques mostly depend on controlled conditions like features and other representations of the human image. Because of significant and uncertain variations of a particular image, it may be a challenging task in gender classification for real-time image processing application, whether it is male, female and others. So that in this document, we propose a Human detection and Face based gender Recognition System (HDFGR); to investigate male or female classification on real life faces using real world face databases. Our proposed approach consists Multi-Scale Invariant Feature Transform (MSIFT) to describes faces and Gaussian distance-based support vector machine (GSVM) classifier is used to classify gender and objects, i.e. male, female and other from features extracted human image datasets. We obtain an experimental performance of 98.7% by applying DSVM with boosted MSIFT features. Our proposed approach gives better classification accuracy and other performance parameters compared to different existing approaches with benchmark and evaluation of possible databases.
Index Terms: Gender Classification, Scale Invariant Feature Transform, Support vector machine, classification accuracy, Adaboost, Feature extraction.

Scope of the Article: