Spatiotemporal Derivative Pattern: A Dynamic Texture Descriptor for Video Matching

Farshid Hajati; Mohammad Tavakolian; Soheila Gheisari; Ajmal Saeed Mian

doi:10.1007/978-3-319-16814-2_41

Spatiotemporal Derivative Pattern: A Dynamic Texture Descriptor for Video Matching

Farshid Hajati
, Mohammad Tavakolian
, Soheila Gheisari
, Ajmal Saeed Mian

Computer Science

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

Abstract

We present Spatiotemporal Derivative Pattern (SDP), a descriptor for dynamic textures. Using local continuous circular and spiral neighborhoods within video segments, SDP encodes the derivatives of the directional spatiotemporal patterns into a binary code. The main strength of SDP is that it uses fewer frames per segment to extract more distinctive features for efficient representation and accurate classification of the dynamic textures. The proposed SDP is tested on the Honda/UCSD and the YouTube face databases for video based face recognition and on the Dynamic Texture database for dynamic texture classification. Comparisons with existing state-of-the-art methods show that the proposed SDP achieves the overall best performance on all three databases. To the best of our knowledge, our algorithm achieves the highest results reported to date on the challenging YouTube face database.

Original language	English
Title of host publication	The ACCV 2014 12th Asian Conference on Computer Vision1-5 November, 2014
Editors	Daniel Cremers, Ian Reid, Hideo Saito, Ming-Hsuan Yang
Place of Publication	Cham, Switzerland
Publisher	Springer International Publishing
Pages	626-641
ISBN (Print)	9783319168135, 9783319168142
DOIs	https://doi.org/10.1007/978-3-319-16814-2_41
Publication status	Published - 17 Apr 2015
Event	The ACCV 2014 12th Asian Conference on Computer Vision - National University of Singapore, Singapore Duration: 1 Nov 2014 → 5 Nov 2014

Conference

Conference	The ACCV 2014 12th Asian Conference on Computer Vision
City	Singapore
Period	1/11/14 → 5/11/14

Access to Document

10.1007/978-3-319-16814-2_41

https://hdl.handle.net/1959.11/70776

Handle.net

Persistent link

Cite this

Hajati, F., Tavakolian, M., Gheisari, S., & Mian, A. S. (2015). Spatiotemporal Derivative Pattern: A Dynamic Texture Descriptor for Video Matching. In D. Cremers, I. Reid, H. Saito, & M.-H. Yang (Eds.), The ACCV 2014 12th Asian Conference on Computer Vision1-5 November, 2014 (pp. 626-641). Springer International Publishing. https://doi.org/10.1007/978-3-319-16814-2_41

@inproceedings{8ef198fdeb81470b8a1f021e4b8994fc,

title = "Spatiotemporal Derivative Pattern: A Dynamic Texture Descriptor for Video Matching",

abstract = "We present Spatiotemporal Derivative Pattern (SDP), a descriptor for dynamic textures. Using local continuous circular and spiral neighborhoods within video segments, SDP encodes the derivatives of the directional spatiotemporal patterns into a binary code. The main strength of SDP is that it uses fewer frames per segment to extract more distinctive features for efficient representation and accurate classification of the dynamic textures. The proposed SDP is tested on the Honda/UCSD and the YouTube face databases for video based face recognition and on the Dynamic Texture database for dynamic texture classification. Comparisons with existing state-of-the-art methods show that the proposed SDP achieves the overall best performance on all three databases. To the best of our knowledge, our algorithm achieves the highest results reported to date on the challenging YouTube face database.",

author = "Farshid Hajati and Mohammad Tavakolian and Soheila Gheisari and Mian, \{Ajmal Saeed\}",

year = "2015",

month = apr,

day = "17",

doi = "10.1007/978-3-319-16814-2\_41",

language = "English",

isbn = "9783319168135",

pages = "626--641",

editor = "Daniel Cremers and Ian Reid and Hideo Saito and Ming-Hsuan Yang",

booktitle = "The ACCV 2014 12th Asian Conference on Computer Vision1-5 November, 2014",

publisher = "Springer International Publishing",

note = "The ACCV 2014 12th Asian Conference on Computer Vision ; Conference date: 01-11-2014 Through 05-11-2014",

}

Hajati, F, Tavakolian, M, Gheisari, S & Mian, AS 2015, Spatiotemporal Derivative Pattern: A Dynamic Texture Descriptor for Video Matching. in D Cremers, I Reid, H Saito & M-H Yang (eds), The ACCV 2014 12th Asian Conference on Computer Vision1-5 November, 2014. Springer International Publishing, Cham, Switzerland, pp. 626-641, The ACCV 2014 12th Asian Conference on Computer Vision, Singapore, 1/11/14. https://doi.org/10.1007/978-3-319-16814-2_41

Spatiotemporal Derivative Pattern: A Dynamic Texture Descriptor for Video Matching. / Hajati, Farshid; Tavakolian, Mohammad; Gheisari, Soheila et al.
The ACCV 2014 12th Asian Conference on Computer Vision1-5 November, 2014. ed. / Daniel Cremers; Ian Reid; Hideo Saito; Ming-Hsuan Yang. Cham, Switzerland: Springer International Publishing, 2015. p. 626-641.

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

TY - GEN

T1 - Spatiotemporal Derivative Pattern

T2 - The ACCV 2014 12th Asian Conference on Computer Vision

AU - Hajati, Farshid

AU - Tavakolian, Mohammad

AU - Gheisari, Soheila

AU - Mian, Ajmal Saeed

PY - 2015/4/17

Y1 - 2015/4/17

N2 - We present Spatiotemporal Derivative Pattern (SDP), a descriptor for dynamic textures. Using local continuous circular and spiral neighborhoods within video segments, SDP encodes the derivatives of the directional spatiotemporal patterns into a binary code. The main strength of SDP is that it uses fewer frames per segment to extract more distinctive features for efficient representation and accurate classification of the dynamic textures. The proposed SDP is tested on the Honda/UCSD and the YouTube face databases for video based face recognition and on the Dynamic Texture database for dynamic texture classification. Comparisons with existing state-of-the-art methods show that the proposed SDP achieves the overall best performance on all three databases. To the best of our knowledge, our algorithm achieves the highest results reported to date on the challenging YouTube face database.

AB - We present Spatiotemporal Derivative Pattern (SDP), a descriptor for dynamic textures. Using local continuous circular and spiral neighborhoods within video segments, SDP encodes the derivatives of the directional spatiotemporal patterns into a binary code. The main strength of SDP is that it uses fewer frames per segment to extract more distinctive features for efficient representation and accurate classification of the dynamic textures. The proposed SDP is tested on the Honda/UCSD and the YouTube face databases for video based face recognition and on the Dynamic Texture database for dynamic texture classification. Comparisons with existing state-of-the-art methods show that the proposed SDP achieves the overall best performance on all three databases. To the best of our knowledge, our algorithm achieves the highest results reported to date on the challenging YouTube face database.

U2 - 10.1007/978-3-319-16814-2_41

DO - 10.1007/978-3-319-16814-2_41

M3 - Conference contribution

SN - 9783319168135

SN - 9783319168142

SP - 626

EP - 641

BT - The ACCV 2014 12th Asian Conference on Computer Vision1-5 November, 2014

A2 - Cremers, Daniel

A2 - Reid, Ian

A2 - Saito, Hideo

A2 - Yang, Ming-Hsuan

PB - Springer International Publishing

CY - Cham, Switzerland

Y2 - 1 November 2014 through 5 November 2014

ER -

Hajati F, Tavakolian M, Gheisari S, Mian AS. Spatiotemporal Derivative Pattern: A Dynamic Texture Descriptor for Video Matching. In Cremers D, Reid I, Saito H, Yang MH, editors, The ACCV 2014 12th Asian Conference on Computer Vision1-5 November, 2014. Cham, Switzerland: Springer International Publishing. 2015. p. 626-641 Epub 2015 Jan 1. doi: 10.1007/978-3-319-16814-2_41

Spatiotemporal Derivative Pattern: A Dynamic Texture Descriptor for Video Matching

Abstract

Conference

Access to Document

Handle.net

Fingerprint

Cite this