Joint classification of actions with matrix completion

Sushma Bomma; Neil M. Robertson

doi:10.1109/ICIP.2015.7351306

Joint classification of actions with matrix completion

Sushma Bomma, Neil M. Robertson

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

4 Citations (Scopus)

Abstract

Action classification is one of the crucial research areas with multitude applications. It has witnessed significant developments over last decade. In this paper, we propose to jointly classify actions from more than a single class using Matrix completion. Matrix-completion methods can handle the deficiencies in data very effectively resulting in improved classification accuracy. Features and labels from data are concatenated to form a big matrix with unknown or missing entries in the place of test data labels. Matrix-completion methods fill up these entries using tools from convex optimization resulting in classification. We show that the proposed method achieves improved performance over the recent works on two human action datasets including most popular Weizmann dataset and recently released and more realistic UCF-101 dataset.

Original language	English
Title of host publication	2015 IEEE International Conference on Image Processing (ICIP)
Subtitle of host publication	Proceedings
Publisher	IEEE
Pages	2766-2770
Number of pages	5
ISBN (Print)	9781479983391
DOIs	https://doi.org/10.1109/ICIP.2015.7351306
Publication status	Published - 2015
Event	22nd IEEE International Conference on Image Processing 2015 - Quebec City, Canada Duration: 27 Sept 2015 → 30 Sept 2015 Conference number: 22

Conference

Conference	22nd IEEE International Conference on Image Processing 2015
Abbreviated title	ICIP 2015
Country/Territory	Canada
City	Quebec City
Period	27/09/15 → 30/09/15

Keywords

Compressed Sensing
Convex Optimization
Human Action Classification/Recognition
Matrix Completion

ASJC Scopus subject areas

Software
Computer Vision and Pattern Recognition
Signal Processing

Access to Document

10.1109/ICIP.2015.7351306

Cite this

@inproceedings{fc185e43c6884ce9b7f4d7dd41881526,

title = "Joint classification of actions with matrix completion",

abstract = "Action classification is one of the crucial research areas with multitude applications. It has witnessed significant developments over last decade. In this paper, we propose to jointly classify actions from more than a single class using Matrix completion. Matrix-completion methods can handle the deficiencies in data very effectively resulting in improved classification accuracy. Features and labels from data are concatenated to form a big matrix with unknown or missing entries in the place of test data labels. Matrix-completion methods fill up these entries using tools from convex optimization resulting in classification. We show that the proposed method achieves improved performance over the recent works on two human action datasets including most popular Weizmann dataset and recently released and more realistic UCF-101 dataset.",

keywords = "Compressed Sensing, Convex Optimization, Human Action Classification/Recognition, Matrix Completion",

author = "Sushma Bomma and Robertson, {Neil M.}",

year = "2015",

doi = "10.1109/ICIP.2015.7351306",

language = "English",

isbn = "9781479983391",

pages = "2766--2770",

booktitle = "2015 IEEE International Conference on Image Processing (ICIP)",

publisher = "IEEE",

address = "United States",

note = "22nd IEEE International Conference on Image Processing 2015, ICIP 2015 ; Conference date: 27-09-2015 Through 30-09-2015",

}

TY - GEN

T1 - Joint classification of actions with matrix completion

AU - Bomma, Sushma

AU - Robertson, Neil M.

N1 - Conference code: 22

PY - 2015

Y1 - 2015

N2 - Action classification is one of the crucial research areas with multitude applications. It has witnessed significant developments over last decade. In this paper, we propose to jointly classify actions from more than a single class using Matrix completion. Matrix-completion methods can handle the deficiencies in data very effectively resulting in improved classification accuracy. Features and labels from data are concatenated to form a big matrix with unknown or missing entries in the place of test data labels. Matrix-completion methods fill up these entries using tools from convex optimization resulting in classification. We show that the proposed method achieves improved performance over the recent works on two human action datasets including most popular Weizmann dataset and recently released and more realistic UCF-101 dataset.

AB - Action classification is one of the crucial research areas with multitude applications. It has witnessed significant developments over last decade. In this paper, we propose to jointly classify actions from more than a single class using Matrix completion. Matrix-completion methods can handle the deficiencies in data very effectively resulting in improved classification accuracy. Features and labels from data are concatenated to form a big matrix with unknown or missing entries in the place of test data labels. Matrix-completion methods fill up these entries using tools from convex optimization resulting in classification. We show that the proposed method achieves improved performance over the recent works on two human action datasets including most popular Weizmann dataset and recently released and more realistic UCF-101 dataset.

KW - Compressed Sensing

KW - Convex Optimization

KW - Human Action Classification/Recognition

KW - Matrix Completion

UR - http://www.scopus.com/inward/record.url?scp=84956630665&partnerID=8YFLogxK

U2 - 10.1109/ICIP.2015.7351306

DO - 10.1109/ICIP.2015.7351306

M3 - Conference contribution

AN - SCOPUS:84956630665

SN - 9781479983391

SP - 2766

EP - 2770

BT - 2015 IEEE International Conference on Image Processing (ICIP)

PB - IEEE

T2 - 22nd IEEE International Conference on Image Processing 2015

Y2 - 27 September 2015 through 30 September 2015

ER -

Joint classification of actions with matrix completion

Abstract

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this