AdaBoost.MRF: Boosted Markov Random Forests and Application to Multilevel Activity Recognition

Activity recognition is an important issue in building intelligent monitoring systems. We address the recognition of multilevel activities in this paper via a conditional Markov random field (MRF), known as the dynamic conditional random field (DCRF). Parameter estimation in general MRFs using maxim...

Full description

Bibliographic Details
Main Authors: Truyen, Tran, Phung, Dinh, Bui, H., Venkatesh, Svetha
Other Authors: A. Fitzgibbon
Format: Conference Paper
Published: IEEE Computer Society Conference Publishing Services 2006
Online Access:http://hdl.handle.net/20.500.11937/16060
_version_ 1848749064978432000
author Truyen, Tran
Phung, Dinh
Bui, H.
Venkatesh, Svetha
author2 A. Fitzgibbon
author_facet A. Fitzgibbon
Truyen, Tran
Phung, Dinh
Bui, H.
Venkatesh, Svetha
author_sort Truyen, Tran
building Curtin Institutional Repository
collection Online Access
description Activity recognition is an important issue in building intelligent monitoring systems. We address the recognition of multilevel activities in this paper via a conditional Markov random field (MRF), known as the dynamic conditional random field (DCRF). Parameter estimation in general MRFs using maximum likelihood is known to be computationally challenging (except for extreme cases), and thus we propose an efficient boosting-based algorithm AdaBoost.MRF for this task. Distinct from most existing work, our algorithm can handle hidden variables (missing labels) and is particularly attractive for smarthouse domains where reliable labels are often sparsely observed. Furthermore, our method works exclusively on trees and thus is guaranteed to converge. We apply the AdaBoost.MRF algorithmto a home video surveillance application and demonstrate its efficacy.
first_indexed 2025-11-14T07:15:00Z
format Conference Paper
id curtin-20.500.11937-16060
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T07:15:00Z
publishDate 2006
publisher IEEE Computer Society Conference Publishing Services
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-160602022-10-27T07:51:49Z AdaBoost.MRF: Boosted Markov Random Forests and Application to Multilevel Activity Recognition Truyen, Tran Phung, Dinh Bui, H. Venkatesh, Svetha A. Fitzgibbon C. Taylor Y. LeCun Activity recognition is an important issue in building intelligent monitoring systems. We address the recognition of multilevel activities in this paper via a conditional Markov random field (MRF), known as the dynamic conditional random field (DCRF). Parameter estimation in general MRFs using maximum likelihood is known to be computationally challenging (except for extreme cases), and thus we propose an efficient boosting-based algorithm AdaBoost.MRF for this task. Distinct from most existing work, our algorithm can handle hidden variables (missing labels) and is particularly attractive for smarthouse domains where reliable labels are often sparsely observed. Furthermore, our method works exclusively on trees and thus is guaranteed to converge. We apply the AdaBoost.MRF algorithmto a home video surveillance application and demonstrate its efficacy. 2006 Conference Paper http://hdl.handle.net/20.500.11937/16060 10.1109/CVPR.2006.49 IEEE Computer Society Conference Publishing Services restricted
spellingShingle Truyen, Tran
Phung, Dinh
Bui, H.
Venkatesh, Svetha
AdaBoost.MRF: Boosted Markov Random Forests and Application to Multilevel Activity Recognition
title AdaBoost.MRF: Boosted Markov Random Forests and Application to Multilevel Activity Recognition
title_full AdaBoost.MRF: Boosted Markov Random Forests and Application to Multilevel Activity Recognition
title_fullStr AdaBoost.MRF: Boosted Markov Random Forests and Application to Multilevel Activity Recognition
title_full_unstemmed AdaBoost.MRF: Boosted Markov Random Forests and Application to Multilevel Activity Recognition
title_short AdaBoost.MRF: Boosted Markov Random Forests and Application to Multilevel Activity Recognition
title_sort adaboost.mrf: boosted markov random forests and application to multilevel activity recognition
url http://hdl.handle.net/20.500.11937/16060