Please use this identifier to cite or link to this item: https://doi.org/10.1145/1291233.1291295
Title: Segregated feedback with performance-based adaptive sampling for interactive news video retrieval
Authors: Luan, H.-B.
Neo, S.-Y. 
Goh, H.-K. 
Zhang, Y.-D.
Lin, S.-X.
Chua, T.-S. 
Keywords: Active learning
News video retrieval
Relevance feedback
Issue Date: 2007
Source: Luan, H.-B.,Neo, S.-Y.,Goh, H.-K.,Zhang, Y.-D.,Lin, S.-X.,Chua, T.-S. (2007). Segregated feedback with performance-based adaptive sampling for interactive news video retrieval. Proceedings of the ACM International Multimedia Conference and Exhibition : 293-296. ScholarBank@NUS Repository. https://doi.org/10.1145/1291233.1291295
Abstract: Existing video research incorporates the use of relevance feedback based on user-dependent interpretations to improve the retrieval results. In this paper, we segregate the process of relevance feedback into 2 distinct facets: (a) recall-directed feedback; and (b) precision-directed feedback. The recall-directed facet employs general features such as text and high level features (HLFs) to maximize efficiency and recall during feedback, making it very suitable for large corpuses. The precision-directed facet on the other hand uses many other multimodal features in an active learning environment for improved accuracy. Combined with a performance-based adaptive sampling strategy, this process continuously re-ranks a subset of instances as the user annotates. Experiments done using TRECVID 2006 dataset show that our approach is efficient and effective. Copyright 2007 ACM.
Source Title: Proceedings of the ACM International Multimedia Conference and Exhibition
URI: http://scholarbank.nus.edu.sg/handle/10635/40595
ISBN: 9781595937025
DOI: 10.1145/1291233.1291295
Appears in Collections:Staff Publications

Show full item record
Files in This Item:
There are no files associated with this item.

SCOPUSTM   
Citations

14
checked on Jan 22, 2018

Page view(s)

51
checked on Jan 19, 2018

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.