Exploring video clips in a vast collection of videos is a difficult task. It is necessary to provide an efficient system for users to express the information need for sought events in that video collection. Thus, we propose to develop AVSeeker – an active video retrieval engine – to assist users in finding appropriate moments in videos with two main query types: textual descriptions and visual examples. The main characteristic of AVSeekeris that we change the retrieval engine from a passive system to an active one, which narrows the search space by gaining clues from users through an interactive relevance feedback manner. The AVSeeker is based on the LifeSeeker system from the annual Lifelog Search Challenge with the addition of an interactive relevance feedback via concept recommendation, enriched temporal concepts, and query-bysketch functionality
Metadata
Item Type:
Conference or Workshop Item (Paper)
Event Type:
Conference
Refereed:
Yes
Uncontrolled Keywords:
Video retrieval system; Interactive relevance feedback retrieval system; Query by sketch
This item is licensed under a Creative Commons Attribution-NonCommercial-Share Alike 3.0 License. View License
Funders:
ADAPT Centre (Grant 13/RC/ 2106; 13/RC/2106 2289 P2), Insight Centre for Data Analytics (Grant SFI/12/RC/ P2), Science Foundation Ireland Research Centres Programme and co funded by the European Regional Development Fund.
ID Code:
27038
Deposited On:
20 Apr 2022 11:37 by
Tu-Khiem Le
. Last Modified 03 Mar 2023 12:46