Audio-Visual Classification of Sports Types

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

3 Citationer (Scopus)
145 Downloads (Pure)

Resumé

In this work we propose a method for classification of sports types from combined audio and visual features extracted from thermal video. From audio Mel Frequency Cepstral Coefficients (MFCC) are extracted, and PCA are
applied to reduce the feature space to 10 dimensions. From the visual modality short trajectories are constructed to represent the motion of players. From these, four motion features are extracted and combined directly with audio features for classification. A k-nearest neighbour classifier is applied for classification of 180 1-minute video sequences from three sports types. Using 10-fold cross validation a correct classification rate of 96.11% is obtained with multimodal features, compared to 86.67% and 90.00% using only visual or audio features, respectively.
OriginalsprogEngelsk
Titel2015 IEEE International Conference on Computer Vision Workshops (ICCVW)
ForlagIEEE
Publikationsdatodec. 2015
Sider768-773
ISBN (Trykt)978-1-4673-8390-5
DOI
StatusUdgivet - dec. 2015
BegivenhedInternational Conference on Computer Vision - Santiago, Chile
Varighed: 12 dec. 201518 dec. 2015

Konference

KonferenceInternational Conference on Computer Vision
LandChile
BySantiago
Periode12/12/201518/12/2015

Fingerprint

Sports
Classifiers
Trajectories

Citer dette

Gade, R., Abou-Zleikha, M., Christensen, M. G., & Moeslund, T. B. (2015). Audio-Visual Classification of Sports Types. I 2015 IEEE International Conference on Computer Vision Workshops (ICCVW) (s. 768-773). IEEE. https://doi.org/10.1109/ICCVW.2015.104
Gade, Rikke ; Abou-Zleikha, Mohamed ; Christensen, Mads Græsbøll ; Moeslund, Thomas B. / Audio-Visual Classification of Sports Types. 2015 IEEE International Conference on Computer Vision Workshops (ICCVW). IEEE, 2015. s. 768-773
@inproceedings{3d8201a8cac64f2ba53338fd33bf62a8,
title = "Audio-Visual Classification of Sports Types",
abstract = "In this work we propose a method for classification of sports types from combined audio and visual features ex- tracted from thermal video. From audio Mel Frequency Cepstral Coefficients (MFCC) are extracted, and PCA are applied to reduce the feature space to 10 dimensions. From the visual modality short trajectories are constructed to rep- resent the motion of players. From these, four motion fea- tures are extracted and combined directly with audio fea- tures for classification. A k-nearest neighbour classifier is applied for classification of 180 1-minute video sequences from three sports types. Using 10-fold cross validation a correct classification rate of 96.11{\%} is obtained with multi- modal features, compared to 86.67{\%} and 90.00{\%} using only visual or audio features, respectively.",
author = "Rikke Gade and Mohamed Abou-Zleikha and Christensen, {Mads Gr{\ae}sb{\o}ll} and Moeslund, {Thomas B.}",
year = "2015",
month = "12",
doi = "10.1109/ICCVW.2015.104",
language = "English",
isbn = "978-1-4673-8390-5",
pages = "768--773",
booktitle = "2015 IEEE International Conference on Computer Vision Workshops (ICCVW)",
publisher = "IEEE",
address = "United States",

}

Gade, R, Abou-Zleikha, M, Christensen, MG & Moeslund, TB 2015, Audio-Visual Classification of Sports Types. i 2015 IEEE International Conference on Computer Vision Workshops (ICCVW). IEEE, s. 768-773, International Conference on Computer Vision, Santiago, Chile, 12/12/2015. https://doi.org/10.1109/ICCVW.2015.104

Audio-Visual Classification of Sports Types. / Gade, Rikke; Abou-Zleikha, Mohamed; Christensen, Mads Græsbøll; Moeslund, Thomas B.

2015 IEEE International Conference on Computer Vision Workshops (ICCVW). IEEE, 2015. s. 768-773.

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

TY - GEN

T1 - Audio-Visual Classification of Sports Types

AU - Gade, Rikke

AU - Abou-Zleikha, Mohamed

AU - Christensen, Mads Græsbøll

AU - Moeslund, Thomas B.

PY - 2015/12

Y1 - 2015/12

N2 - In this work we propose a method for classification of sports types from combined audio and visual features ex- tracted from thermal video. From audio Mel Frequency Cepstral Coefficients (MFCC) are extracted, and PCA are applied to reduce the feature space to 10 dimensions. From the visual modality short trajectories are constructed to rep- resent the motion of players. From these, four motion fea- tures are extracted and combined directly with audio fea- tures for classification. A k-nearest neighbour classifier is applied for classification of 180 1-minute video sequences from three sports types. Using 10-fold cross validation a correct classification rate of 96.11% is obtained with multi- modal features, compared to 86.67% and 90.00% using only visual or audio features, respectively.

AB - In this work we propose a method for classification of sports types from combined audio and visual features ex- tracted from thermal video. From audio Mel Frequency Cepstral Coefficients (MFCC) are extracted, and PCA are applied to reduce the feature space to 10 dimensions. From the visual modality short trajectories are constructed to rep- resent the motion of players. From these, four motion fea- tures are extracted and combined directly with audio fea- tures for classification. A k-nearest neighbour classifier is applied for classification of 180 1-minute video sequences from three sports types. Using 10-fold cross validation a correct classification rate of 96.11% is obtained with multi- modal features, compared to 86.67% and 90.00% using only visual or audio features, respectively.

U2 - 10.1109/ICCVW.2015.104

DO - 10.1109/ICCVW.2015.104

M3 - Article in proceeding

SN - 978-1-4673-8390-5

SP - 768

EP - 773

BT - 2015 IEEE International Conference on Computer Vision Workshops (ICCVW)

PB - IEEE

ER -

Gade R, Abou-Zleikha M, Christensen MG, Moeslund TB. Audio-Visual Classification of Sports Types. I 2015 IEEE International Conference on Computer Vision Workshops (ICCVW). IEEE. 2015. s. 768-773 https://doi.org/10.1109/ICCVW.2015.104