Face Annotation Weakly Labelled Images using Nearest Neighbour Calculation

Author(s):

Ann Veena Paul , M.A.College of Engineering and Technology,Kothamangalam,Kerala; Merin K Kurian, M.A.College of Engineering and Technology,Kothamangalam,Kerala; Eldo P Ealias, M.A.College of Engineering and Technology,Kothamangalam,Kerala

Keywords:

Locality Sensitive Hashing, Label Refinement, Face Annotation, Clustering Based Approximation, Performance

Abstract:

Social networks such as face book, twitter etc. are widely used in our day to day life. Face image retrieval using content based method is an emerging technology in many real world applications for automatic face annotation. Our main goal is to retrieve the similar images from large scale database using content based. One challenging problem for search-based face annotation scheme is how to effectively perform annotation by exploiting the list of most similar facial images and their weak labels that are often noisy and incomplete. To tackle this problem, we propose an effective label refinement approach based on the semantic closeness of the image with name-list with which we had crawled the web search engines. Locality-Sensitive Hashing (LSH) to index the facial features in our solution and a Clustering of images on name-level is proposed to improve the efficiency of the annotation. We conduct extensive empirical studies on several web facial image databases to evaluate the proposed classification algorithm from different aspects.

Other Details:

Manuscript Id	:	IJSTEV3I1046
Published in	:	Volume : 3, Issue : 1
Publication Date	:	01/08/2016
Page(s)	:	488-493

Download Article