Searching visual instances with topology checking and context modeling
Instance Search (INS) is a realistic problem initiated by TRECVID, which is to retrieve all occurrences of the querying object, location, or person from a large video collection. It is a fundamental problem with many applications, and also a challenging problem different from the traditional concept...
Saved in:
Main Authors: | , |
---|---|
Format: | text |
Language: | English |
Published: |
Institutional Knowledge at Singapore Management University
2013
|
Subjects: | |
Online Access: | https://ink.library.smu.edu.sg/sis_research/6473 https://ink.library.smu.edu.sg/context/sis_research/article/7476/viewcontent/2461466.2461477.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Singapore Management University |
Language: | English |
id |
sg-smu-ink.sis_research-7476 |
---|---|
record_format |
dspace |
spelling |
sg-smu-ink.sis_research-74762022-01-10T05:43:05Z Searching visual instances with topology checking and context modeling ZHANG, Wei NGO, Chong-wah Instance Search (INS) is a realistic problem initiated by TRECVID, which is to retrieve all occurrences of the querying object, location, or person from a large video collection. It is a fundamental problem with many applications, and also a challenging problem different from the traditional concept or near-duplicate (ND) search, since the relevancy is defined at instance level. True responses could exhibit various visual variations, such as being small on the image with different background, or showing a non-homography spatial configuration. Based on the Bag-of-Words model, we propose two techniques tailored for Instance Search. Specifically, we explore the use of (1) an elastic spatial topology checking technique based on Delaunay Triangulation (DT), and (2) a practical background context modeling method by simulating the “stare” behavior of human eyes. With DT, we improve the quality of visual matching by accumulating evidence from local topology-preserving patches, significantly boosting the ranks of topology consistent results. On the other hand, we increase the information quantity for visual matching with the“stare”model, such that instances appearing in both similar and different background can be highly ranked as results. The proposed techniques are evaluated on the INS datasets of TRECVID, achieving large performance gain with small computation overhead, compared with several existing methods. 2013-04-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/6473 info:doi/10.1145/2461466.2461477 https://ink.library.smu.edu.sg/context/sis_research/article/7476/viewcontent/2461466.2461477.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University context modeling instance search spatial topology checking TRECVID Data Storage Systems Graphics and Human Computer Interfaces |
institution |
Singapore Management University |
building |
SMU Libraries |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
SMU Libraries |
collection |
InK@SMU |
language |
English |
topic |
context modeling instance search spatial topology checking TRECVID Data Storage Systems Graphics and Human Computer Interfaces |
spellingShingle |
context modeling instance search spatial topology checking TRECVID Data Storage Systems Graphics and Human Computer Interfaces ZHANG, Wei NGO, Chong-wah Searching visual instances with topology checking and context modeling |
description |
Instance Search (INS) is a realistic problem initiated by TRECVID, which is to retrieve all occurrences of the querying object, location, or person from a large video collection. It is a fundamental problem with many applications, and also a challenging problem different from the traditional concept or near-duplicate (ND) search, since the relevancy is defined at instance level. True responses could exhibit various visual variations, such as being small on the image with different background, or showing a non-homography spatial configuration. Based on the Bag-of-Words model, we propose two techniques tailored for Instance Search. Specifically, we explore the use of (1) an elastic spatial topology checking technique based on Delaunay Triangulation (DT), and (2) a practical background context modeling method by simulating the “stare” behavior of human eyes. With DT, we improve the quality of visual matching by accumulating evidence from local topology-preserving patches, significantly boosting the ranks of topology consistent results. On the other hand, we increase the information quantity for visual matching with the“stare”model, such that instances appearing in both similar and different background can be highly ranked as results. The proposed techniques are evaluated on the INS datasets of TRECVID, achieving large performance gain with small computation overhead, compared with several existing methods. |
format |
text |
author |
ZHANG, Wei NGO, Chong-wah |
author_facet |
ZHANG, Wei NGO, Chong-wah |
author_sort |
ZHANG, Wei |
title |
Searching visual instances with topology checking and context modeling |
title_short |
Searching visual instances with topology checking and context modeling |
title_full |
Searching visual instances with topology checking and context modeling |
title_fullStr |
Searching visual instances with topology checking and context modeling |
title_full_unstemmed |
Searching visual instances with topology checking and context modeling |
title_sort |
searching visual instances with topology checking and context modeling |
publisher |
Institutional Knowledge at Singapore Management University |
publishDate |
2013 |
url |
https://ink.library.smu.edu.sg/sis_research/6473 https://ink.library.smu.edu.sg/context/sis_research/article/7476/viewcontent/2461466.2461477.pdf |
_version_ |
1770575970242134016 |