Fast action detection via discriminative random forest voting and top-K subvolume search
Multiclass action detection in complex scenes is a challenging problem because of cluttered backgrounds and the large intra-class variations in each type of actions. To achieve efficient and robust action detection, we characterize a video as a collection of spatio-temporal interest points, and loca...
Guardado en:
Publicado: |
2011
|
---|---|
Materias: | |
Acceso en línea: | https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_15209210_v13_n3_p507_Yu http://hdl.handle.net/20.500.12110/paper_15209210_v13_n3_p507_Yu |
Aporte de: |
id |
paper:paper_15209210_v13_n3_p507_Yu |
---|---|
record_format |
dspace |
spelling |
paper:paper_15209210_v13_n3_p507_Yu2023-06-08T16:19:15Z Fast action detection via discriminative random forest voting and top-K subvolume search Action detection branch and bound random forest top-K search Action detection branch and bound Complex scenes Data sets Detection methods Interest points Intra-class variation Multi-class Mutual informations Orders of magnitude random forest Random forests Search Algorithms Spatio-temporal Subvolumes top-K search Linear programming Decision trees Multiclass action detection in complex scenes is a challenging problem because of cluttered backgrounds and the large intra-class variations in each type of actions. To achieve efficient and robust action detection, we characterize a video as a collection of spatio-temporal interest points, and locate actions via finding spatio-temporal video subvolumes of the highest mutual information score towards each action class. A random forest is constructed to efficiently generate discriminative votes from individual interest points, and a fast top-K subvolume search algorithm is developed to find all action instances in a single round of search. Without significantly degrading the performance, such a top-K search can be performed on down-sampled score volumes for more efficient localization. Experiments on a challenging MSR Action Dataset II validate the effectiveness of our proposed multiclass action detection method. The detection speed is several orders of magnitude faster than existing methods. © 2011 IEEE. 2011 https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_15209210_v13_n3_p507_Yu http://hdl.handle.net/20.500.12110/paper_15209210_v13_n3_p507_Yu |
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 detection branch and bound random forest top-K search Action detection branch and bound Complex scenes Data sets Detection methods Interest points Intra-class variation Multi-class Mutual informations Orders of magnitude random forest Random forests Search Algorithms Spatio-temporal Subvolumes top-K search Linear programming Decision trees |
spellingShingle |
Action detection branch and bound random forest top-K search Action detection branch and bound Complex scenes Data sets Detection methods Interest points Intra-class variation Multi-class Mutual informations Orders of magnitude random forest Random forests Search Algorithms Spatio-temporal Subvolumes top-K search Linear programming Decision trees Fast action detection via discriminative random forest voting and top-K subvolume search |
topic_facet |
Action detection branch and bound random forest top-K search Action detection branch and bound Complex scenes Data sets Detection methods Interest points Intra-class variation Multi-class Mutual informations Orders of magnitude random forest Random forests Search Algorithms Spatio-temporal Subvolumes top-K search Linear programming Decision trees |
description |
Multiclass action detection in complex scenes is a challenging problem because of cluttered backgrounds and the large intra-class variations in each type of actions. To achieve efficient and robust action detection, we characterize a video as a collection of spatio-temporal interest points, and locate actions via finding spatio-temporal video subvolumes of the highest mutual information score towards each action class. A random forest is constructed to efficiently generate discriminative votes from individual interest points, and a fast top-K subvolume search algorithm is developed to find all action instances in a single round of search. Without significantly degrading the performance, such a top-K search can be performed on down-sampled score volumes for more efficient localization. Experiments on a challenging MSR Action Dataset II validate the effectiveness of our proposed multiclass action detection method. The detection speed is several orders of magnitude faster than existing methods. © 2011 IEEE. |
title |
Fast action detection via discriminative random forest voting and top-K subvolume search |
title_short |
Fast action detection via discriminative random forest voting and top-K subvolume search |
title_full |
Fast action detection via discriminative random forest voting and top-K subvolume search |
title_fullStr |
Fast action detection via discriminative random forest voting and top-K subvolume search |
title_full_unstemmed |
Fast action detection via discriminative random forest voting and top-K subvolume search |
title_sort |
fast action detection via discriminative random forest voting and top-k subvolume search |
publishDate |
2011 |
url |
https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_15209210_v13_n3_p507_Yu http://hdl.handle.net/20.500.12110/paper_15209210_v13_n3_p507_Yu |
_version_ |
1768544100456333312 |