High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

A Novel Approach of Fuzzy Based Semi-Automatic Annotation for Similar Facial Images

Author(s):

A. Nithya , NGM college; K. Haridas, NGM college

Keywords:

Face annotation, weak label, optimization, ULR algorithms, fuzzy

Abstract

Auto face annotation is an important role for many real-world multimedia applications. Recently search based face annotation paradigm is one of the research challenge in computer vision and image analysis. In this paper, we present the problem is to annotate the most weakly labeled facial images are duplicate names, noisy and incomplete. To handle this problem we used an effective semi-automatic annotation methodology with unsupervised label refinement (ULR) approach for refining the labels in facial images by using some machine learning techniques and fuzzy clustering-based approximation algorithm is used to improve the scalability considerably. Finally to develop an optimization algorithm for solving a large scale learning task. The result of this proposed ULR algorithm can improve the performance than other ULR algorithms in weak label matrix.

Other Details

Paper ID: IJSRDV2I7209
Published in: Volume : 2, Issue : 7
Publication Date: 01/10/2014
Page(s): 590-593

Article Preview

Download Article