Efficient search of Top-K video subvolumes for multi-instance action detection
Action detection was formulated as a subvolume mutual information maximization problem in [8], where each subvolume identifies where and when the action occurs in the video. Despite the fact that the proposed branch-and-bound algorithm can find the best subvolume efficiently for low resolution video...
Guardado en:
Autores principales: | , , |
---|---|
Formato: | CONF |
Materias: | |
Acceso en línea: | http://hdl.handle.net/20.500.12110/paper_97814244_v_n_p328_Goussies |
Aporte de: |
id |
todo:paper_97814244_v_n_p328_Goussies |
---|---|
record_format |
dspace |
spelling |
todo:paper_97814244_v_n_p328_Goussies2023-10-03T16:43:10Z Efficient search of Top-K video subvolumes for multi-instance action detection Goussies, N.A. Liu, Z. Yuan, J. Action recognition Branch-and-bound Action recognition Branch and bounds Branch-and-bound algorithms Data sets High resolution High spatial resolution Low resolution video Mutual information maximization Search technique Spatial resolution Spatial scale Subvolumes Upper bound Video sequences Image resolution Video recording Algorithms Action detection was formulated as a subvolume mutual information maximization problem in [8], where each subvolume identifies where and when the action occurs in the video. Despite the fact that the proposed branch-and-bound algorithm can find the best subvolume efficiently for low resolution videos, it is still not efficient enough to perform multiinstance detection in videos of high spatial resolution. In this paper we develop an algorithm that further speeds up the subvolume search and targets on real-time multi-instance action detection for high resolution videos (e.g. 320 × 240 or higher). Unlike the previous branch-and-bound search technique which restarts a new search for each action instance, we find the Top-K subvolumes simultaneously with a single round of search. To handle the larger spatial resolution, we downsample the volume of videos for a more efficient upperbound estimation. To validate our algorithm, we perform experiments on a challenging dataset of 54 video sequences where each video consists of several actions performed by different people in a crowded environment. The experiments show that our method is not only efficient, but also capable of handling action variations caused by performing speed and style changes, spatial scale changes, as well as cluttered and moving background. © 2010 IEEE. CONF info:eu-repo/semantics/openAccess http://creativecommons.org/licenses/by/2.5/ar http://hdl.handle.net/20.500.12110/paper_97814244_v_n_p328_Goussies |
institution |
Universidad de Buenos Aires |
institution_str |
I-28 |
repository_str |
R-134 |
collection |
Biblioteca Digital - Facultad de Ciencias Exactas y Naturales (UBA) |
topic |
Action recognition Branch-and-bound Action recognition Branch and bounds Branch-and-bound algorithms Data sets High resolution High spatial resolution Low resolution video Mutual information maximization Search technique Spatial resolution Spatial scale Subvolumes Upper bound Video sequences Image resolution Video recording Algorithms |
spellingShingle |
Action recognition Branch-and-bound Action recognition Branch and bounds Branch-and-bound algorithms Data sets High resolution High spatial resolution Low resolution video Mutual information maximization Search technique Spatial resolution Spatial scale Subvolumes Upper bound Video sequences Image resolution Video recording Algorithms Goussies, N.A. Liu, Z. Yuan, J. Efficient search of Top-K video subvolumes for multi-instance action detection |
topic_facet |
Action recognition Branch-and-bound Action recognition Branch and bounds Branch-and-bound algorithms Data sets High resolution High spatial resolution Low resolution video Mutual information maximization Search technique Spatial resolution Spatial scale Subvolumes Upper bound Video sequences Image resolution Video recording Algorithms |
description |
Action detection was formulated as a subvolume mutual information maximization problem in [8], where each subvolume identifies where and when the action occurs in the video. Despite the fact that the proposed branch-and-bound algorithm can find the best subvolume efficiently for low resolution videos, it is still not efficient enough to perform multiinstance detection in videos of high spatial resolution. In this paper we develop an algorithm that further speeds up the subvolume search and targets on real-time multi-instance action detection for high resolution videos (e.g. 320 × 240 or higher). Unlike the previous branch-and-bound search technique which restarts a new search for each action instance, we find the Top-K subvolumes simultaneously with a single round of search. To handle the larger spatial resolution, we downsample the volume of videos for a more efficient upperbound estimation. To validate our algorithm, we perform experiments on a challenging dataset of 54 video sequences where each video consists of several actions performed by different people in a crowded environment. The experiments show that our method is not only efficient, but also capable of handling action variations caused by performing speed and style changes, spatial scale changes, as well as cluttered and moving background. © 2010 IEEE. |
format |
CONF |
author |
Goussies, N.A. Liu, Z. Yuan, J. |
author_facet |
Goussies, N.A. Liu, Z. Yuan, J. |
author_sort |
Goussies, N.A. |
title |
Efficient search of Top-K video subvolumes for multi-instance action detection |
title_short |
Efficient search of Top-K video subvolumes for multi-instance action detection |
title_full |
Efficient search of Top-K video subvolumes for multi-instance action detection |
title_fullStr |
Efficient search of Top-K video subvolumes for multi-instance action detection |
title_full_unstemmed |
Efficient search of Top-K video subvolumes for multi-instance action detection |
title_sort |
efficient search of top-k video subvolumes for multi-instance action detection |
url |
http://hdl.handle.net/20.500.12110/paper_97814244_v_n_p328_Goussies |
work_keys_str_mv |
AT goussiesna efficientsearchoftopkvideosubvolumesformultiinstanceactiondetection AT liuz efficientsearchoftopkvideosubvolumesformultiinstanceactiondetection AT yuanj efficientsearchoftopkvideosubvolumesformultiinstanceactiondetection |
_version_ |
1782023990663970816 |