Human action capturing and classification
In this report, a vision-based framework is proposed for learning and inferring occupant activities at different levels. These levels range from short temporal interval movements, to intermediate level events and long temporal term activities. Our research is focused on using a combined tracking-cla...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Theses and Dissertations |
Language: | English |
Published: |
2008
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/13599 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-13599 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-135992023-03-04T00:39:50Z Human action capturing and classification Feng, Zhou Cham Tat Jen School of Computer Engineering DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision In this report, a vision-based framework is proposed for learning and inferring occupant activities at different levels. These levels range from short temporal interval movements, to intermediate level events and long temporal term activities. Our research is focused on using a combined tracking-classification framework for the unsupervised classification of human action. An initial comparative study was done to evaluate several existing foreground segmentation methods that employ background modeling. Our own probabilistic foreground-background segmentation method is proposed to extract human-centric reference frames. Based on the human-centric reference frames, a principled analysis of the correspondence problem leads to a novel probabilistic action representation called the correspondence-ambiguous feature histogram array (CAFHA) that is robust to variations across similar actions. CAFHA is shown to be effective in unsupervised action classification and quasi real-time action inference. A novel feature selection method is proposed to select the optimal features to improve the CAFHA representation, such that the best discrimination between different action clusters may be found via unsupervised spectral clustering. Finally, a number of potential future directions are proposed that are targeted at further improvements to our framework and creating new research methods required to recognize human activity at longer temporal scales. MASTER OF ENGINEERING (SCE) 2008-06-17T02:55:55Z 2008-10-20T09:58:09Z 2008-06-17T02:55:55Z 2008-10-20T09:58:09Z 2006 2006 Thesis Feng, Z. (2006). Human action capturing and classification. Master’s thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/13599 10.32657/10356/13599 en 113 p. application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision |
spellingShingle |
DRNTU::Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision Feng, Zhou Human action capturing and classification |
description |
In this report, a vision-based framework is proposed for learning and inferring occupant activities at different levels. These levels range from short temporal interval movements, to intermediate level events and long temporal term activities. Our research is focused on using a combined tracking-classification framework for the unsupervised classification of human action. An initial comparative study was done to evaluate several existing foreground segmentation methods that employ background modeling. Our own probabilistic foreground-background segmentation method is proposed to extract human-centric reference frames. Based on the human-centric reference frames, a principled analysis of the correspondence problem leads to a novel probabilistic action representation called the correspondence-ambiguous feature histogram array (CAFHA) that is robust to variations across similar actions. CAFHA is shown to be effective in unsupervised action classification and quasi real-time action inference. A novel feature selection method is proposed to select the optimal features to improve the CAFHA representation, such that the best discrimination between different action clusters may be found via unsupervised spectral clustering. Finally, a number of potential future directions are proposed that are targeted at further improvements to our framework and creating new research methods required to recognize human activity at longer temporal scales. |
author2 |
Cham Tat Jen |
author_facet |
Cham Tat Jen Feng, Zhou |
format |
Theses and Dissertations |
author |
Feng, Zhou |
author_sort |
Feng, Zhou |
title |
Human action capturing and classification |
title_short |
Human action capturing and classification |
title_full |
Human action capturing and classification |
title_fullStr |
Human action capturing and classification |
title_full_unstemmed |
Human action capturing and classification |
title_sort |
human action capturing and classification |
publishDate |
2008 |
url |
https://hdl.handle.net/10356/13599 |
_version_ |
1759857463771791360 |