Harvesting collective Images for Bi-Concept exploration
Journal Title: International Journal of Advanced Research in Computer Engineering & Technology(IJARCET) - Year 2013, Vol 2, Issue 4
Abstract
Noised positive as well as instructive pessimistic research examples commencing the communal web, to become skilled at bi-concept detectors beginning these examples, and to apply them in a search steam engine for retrieve bi-concepts in unlabeled metaphors. We study the activities of our bi-concept search engine using 1.2 M social-tagged metaphors as a data source. Our experiments designate that harvest examples for bi-concepts differs from inveterate single-concept method, yet the examples can be composed with high accurateness using a multi-modal approach. We find that straight erudition bi-concepts is superior than oracle linear fusion of single-concept detectors Searching for the co-occurrence of two visual concepts in unlabeled images is an important step towards answering composite user queries. Traditional illustration search methods use combinations of the confidence scores of individual concept detectors to tackle such queries. Here introduce the belief of bi-concepts, an innovative concept-based retrieval method that is straightforwardly the examples can be composed with high accuracy using a multi-modal approach. A multimodal approach is that for getting metadata and visual features is used to gather many high-quality images from the Web. And also we consider other factors are bag of words, collection information from the social sites. As we said above methods, we will do training session. First, the images are re ranked based on the text surrounding the image and metadata features. An integer of methods is compared for this re position. Second, the top-ranked metaphors are used as (label) training images and VF and bag of words is learned to improve the ranking further and finally we will get bi-Concepts images. In addition, a scene with two concepts present tends to be visually more composite, requiring multi-modal analysis. Given these difficulties, effective bi-concept search demands an approach to harvesting appropriate examples from social-tagged images for learning bi-concept detectors. We present a multi-modal approach to collect de-noised positive as well as informative negative training examples from the social web. We learn bi-concept detectors from these examples and in a while apply them for retrieving bi-concepts in unlabeled images.
Authors and Affiliations
B. Nithya priya , K . P. kaliyamurthie
Keywords
Related Articles
- EP ID EP156951
- DOI -
- Views 126
- Downloads 0
Image Denoising Using Curvelet Transform Using Log Gabor Filter
— In this paper we propose a new method to reduce noise in digital image. Images corrupted by Gaussian Noise is still a classical problem. To reduce the noise or to improve the quality of image we have used two par...
Optical Character Recognition Techniques: A survey
A Java Applet Based Tool for Histogram Equalization and Edge Detection
Edge detection is a fundamentally important operation in image processing. It is the foundation of feature extraction and image information interpretation. It gives a description of the curvature of objects in an...
An Efficient Authentication and Access Control Scheme to Protect Integrity of Anonymous Networks
Concerns about privacy and security have received greater attention with the rapid growth and public acceptance of the Internet and the pervasive deployment of various network technologies. Anonymous network servic...
SEMANTIC BASED LEARNING APPROACH IN EDUCATION SYSTEM
Extensive Research works in the field of knowledge base Learning are represented by a broad spectrum of applications .Education system transform from class room system to E-Learning system which are ranged from virtual c...