English  |  正體中文  |  简体中文  |  Items with full text/Total items : 27228/39071
Visitors : 2413715      Online Users : 67
RC Version 4.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Adv. Search
LoginUploadHelpAboutAdminister

Please use this identifier to cite or link to this item: http://ntour.ntou.edu.tw:8080/ir/handle/987654321/28016

Title: Learning Visual Concepts from Image Instances
Authors: Jun-Wei Hsieh;Cheng-Chin Chiang;Yea-Shuan Huang;W. E. L. Grimson
Contributors: NTOU:Department of Computer Science and Engineering
國立臺灣海洋大學:資訊工程學系
Keywords: multiple instances;diverse density algorithm;relevance feedback;region instances;image retrieval
Date: 2004-11
Issue Date: 2011-10-21T02:35:10Z
Publisher: Journal of Information Science and Engineering
Abstract: Abstract:This paper presents a novel method to retrieve images by learning the commonality of instances from a set of training examples. The proposed scheme uses a coarse-to-fine algorithm (learner) to find the desired visual concepts from a set of instances (extracted from the above trained examples) for successful image retrieval. The learner at the coarse stage attempts to partition training data into two smaller compact sets ( relevant and irrelevant) for reducing the size of training examples and thus improving the efficiency of concept learning at the refined stage. At the refined stage, a verification scheme is proposed to verify each instance in the selected compact set by examining its indexing and filtering capabilities based on a pool of images. This verification process is very different to other learning methods which consider “learning” only on the user-provided samples and don’t check the above retrieving and filtering capabilities of the selected concept. Due to the extra consideration, the wanted visual concept can be learned more accurately and leads to significant improvements in image retrieval. Since no time-consuming optimization process is involved, all the desired visual concepts can be learned on line and adapted to user’s different requests. Experimental results are provided to verify the superiority of the proposed method.
Relation: 20(6), pp.1197-1212
URI: http://ntour.ntou.edu.tw/handle/987654321/28016
Appears in Collections:[資訊工程學系] 期刊論文

Files in This Item:

File Description SizeFormat
index.html0KbHTML319View/Open


All items in NTOUR are protected by copyright, with all rights reserved.

 


著作權政策宣告: 本網站之內容為國立臺灣海洋大學所收錄之機構典藏,無償提供學術研究與公眾教育等公益性使用,請合理使用本網站之內容,以尊重著作權人之權益。
網站維護: 海大圖資處 圖書系統組
DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - Feedback