Home > Research > Publications & Outputs > Action Recognition Using 3D histograms of Textu...

Research

Associated organisational unit

Links

Text available via DOI:

View graph of relations

Action Recognition Using 3D histograms of Texture and A Multi-class Boosting Classifier

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Published

Standard

Action Recognition Using 3D histograms of Texture and A Multi-class Boosting Classifier. / Zhang, Baochang; Yang, Yu; Chen, Chen et al.
In: IEEE Transactions on Image Processing, Vol. 26, No. 10, 10.2017, p. 4648-4660.

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Harvard

Zhang, B, Yang, Y, Chen, C, Yang, L, Han, J & Shao, L 2017, 'Action Recognition Using 3D histograms of Texture and A Multi-class Boosting Classifier', IEEE Transactions on Image Processing, vol. 26, no. 10, pp. 4648-4660. https://doi.org/10.1109/TIP.2017.2718189

APA

Zhang, B., Yang, Y., Chen, C., Yang, L., Han, J., & Shao, L. (2017). Action Recognition Using 3D histograms of Texture and A Multi-class Boosting Classifier. IEEE Transactions on Image Processing, 26(10), 4648-4660. https://doi.org/10.1109/TIP.2017.2718189

Vancouver

Zhang B, Yang Y, Chen C, Yang L, Han J, Shao L. Action Recognition Using 3D histograms of Texture and A Multi-class Boosting Classifier. IEEE Transactions on Image Processing. 2017 Oct;26(10):4648-4660. Epub 2017 Jun 21. doi: 10.1109/TIP.2017.2718189

Author

Zhang, Baochang ; Yang, Yu ; Chen, Chen et al. / Action Recognition Using 3D histograms of Texture and A Multi-class Boosting Classifier. In: IEEE Transactions on Image Processing. 2017 ; Vol. 26, No. 10. pp. 4648-4660.

Bibtex

@article{9fbe028531414c6889004f6ba7a361d6,
title = "Action Recognition Using 3D histograms of Texture and A Multi-class Boosting Classifier",
abstract = "Human action recognition is an important yet challenging task. This paper presents a low-cost descriptor called 3D histograms of texture (3DHoTs) to extract discriminant features from a sequence of depth maps. 3DHoTs are derived from projecting depth frames onto three orthogonal Cartesian planes, i.e., the frontal, side, and top planes, and thus compactly characterize the salient information of a specific action, on which texture features are calculated to represent the action. Besides this fast feature descriptor, a new multi-class boosting classifier (MBC) is also proposed to efficiently exploit different kinds of features in a unified framework for action classification. Compared with the existing boosting frameworks, we add a new multi-class constraint into the objective function, which helps to maintain a better margin distribution by maximizing the mean of margin, whereas still minimizing the variance of margin. Experiments on the MSRAction3D, MSRGesture3D, MSRActivity3D, and UTD-MHAD data sets demonstrate that the proposed system combining 3DHoTs and MBC is superior to the state of the art.",
author = "Baochang Zhang and Yu Yang and Chen Chen and Linlin Yang and Jungong Han and Ling Shao",
year = "2017",
month = oct,
doi = "10.1109/TIP.2017.2718189",
language = "English",
volume = "26",
pages = "4648--4660",
journal = "IEEE Transactions on Image Processing",
issn = "1057-7149",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "10",

}

RIS

TY - JOUR

T1 - Action Recognition Using 3D histograms of Texture and A Multi-class Boosting Classifier

AU - Zhang, Baochang

AU - Yang, Yu

AU - Chen, Chen

AU - Yang, Linlin

AU - Han, Jungong

AU - Shao, Ling

PY - 2017/10

Y1 - 2017/10

N2 - Human action recognition is an important yet challenging task. This paper presents a low-cost descriptor called 3D histograms of texture (3DHoTs) to extract discriminant features from a sequence of depth maps. 3DHoTs are derived from projecting depth frames onto three orthogonal Cartesian planes, i.e., the frontal, side, and top planes, and thus compactly characterize the salient information of a specific action, on which texture features are calculated to represent the action. Besides this fast feature descriptor, a new multi-class boosting classifier (MBC) is also proposed to efficiently exploit different kinds of features in a unified framework for action classification. Compared with the existing boosting frameworks, we add a new multi-class constraint into the objective function, which helps to maintain a better margin distribution by maximizing the mean of margin, whereas still minimizing the variance of margin. Experiments on the MSRAction3D, MSRGesture3D, MSRActivity3D, and UTD-MHAD data sets demonstrate that the proposed system combining 3DHoTs and MBC is superior to the state of the art.

AB - Human action recognition is an important yet challenging task. This paper presents a low-cost descriptor called 3D histograms of texture (3DHoTs) to extract discriminant features from a sequence of depth maps. 3DHoTs are derived from projecting depth frames onto three orthogonal Cartesian planes, i.e., the frontal, side, and top planes, and thus compactly characterize the salient information of a specific action, on which texture features are calculated to represent the action. Besides this fast feature descriptor, a new multi-class boosting classifier (MBC) is also proposed to efficiently exploit different kinds of features in a unified framework for action classification. Compared with the existing boosting frameworks, we add a new multi-class constraint into the objective function, which helps to maintain a better margin distribution by maximizing the mean of margin, whereas still minimizing the variance of margin. Experiments on the MSRAction3D, MSRGesture3D, MSRActivity3D, and UTD-MHAD data sets demonstrate that the proposed system combining 3DHoTs and MBC is superior to the state of the art.

U2 - 10.1109/TIP.2017.2718189

DO - 10.1109/TIP.2017.2718189

M3 - Journal article

VL - 26

SP - 4648

EP - 4660

JO - IEEE Transactions on Image Processing

JF - IEEE Transactions on Image Processing

SN - 1057-7149

IS - 10

ER -