Sorry, you need to enable JavaScript to visit this website.

Pattern recognition and classification (MLR-PATT)

RANDOM MATRICES MEET MACHINE LEARNING: A LARGE DIMENSIONAL ANALYSIS OF LS-SVM


This article proposes a performance analysis of kernel least squares support vector machines (LS-SVMs) based on a random matrix approach, in the regime where both the dimension of data p and their number n grow large at the same rate. Under a two-class Gaussian mixture model for the input data, we prove that the LS-SVM decision function is asymptotically normal with means and covariances shown to depend explicitly on the derivatives of the kernel function. This provides improved understanding along with new insights into the internal workings of SVM-type methods for large datasets.

Paper Details

Authors:
Zhenyu Liao, Romain Couillet
Submitted On:
9 March 2017 - 6:36pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

RMT4LSSVM-ICASSP.pdf

(32 downloads)

Keywords

Subscribe

[1] Zhenyu Liao, Romain Couillet, "RANDOM MATRICES MEET MACHINE LEARNING: A LARGE DIMENSIONAL ANALYSIS OF LS-SVM", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1731. Accessed: Jun. 28, 2017.
@article{1731-17,
url = {http://sigport.org/1731},
author = {Zhenyu Liao; Romain Couillet },
publisher = {IEEE SigPort},
title = {RANDOM MATRICES MEET MACHINE LEARNING: A LARGE DIMENSIONAL ANALYSIS OF LS-SVM},
year = {2017} }
TY - EJOUR
T1 - RANDOM MATRICES MEET MACHINE LEARNING: A LARGE DIMENSIONAL ANALYSIS OF LS-SVM
AU - Zhenyu Liao; Romain Couillet
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1731
ER -
Zhenyu Liao, Romain Couillet. (2017). RANDOM MATRICES MEET MACHINE LEARNING: A LARGE DIMENSIONAL ANALYSIS OF LS-SVM. IEEE SigPort. http://sigport.org/1731
Zhenyu Liao, Romain Couillet, 2017. RANDOM MATRICES MEET MACHINE LEARNING: A LARGE DIMENSIONAL ANALYSIS OF LS-SVM. Available at: http://sigport.org/1731.
Zhenyu Liao, Romain Couillet. (2017). "RANDOM MATRICES MEET MACHINE LEARNING: A LARGE DIMENSIONAL ANALYSIS OF LS-SVM." Web.
1. Zhenyu Liao, Romain Couillet. RANDOM MATRICES MEET MACHINE LEARNING: A LARGE DIMENSIONAL ANALYSIS OF LS-SVM [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1731

End to end spoofing detection with raw wave CLDNNs

Paper Details

Authors:
Submitted On:
6 March 2017 - 10:03am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICASSP_2017_E2ESpoof_Heinrich_Dinkel.pdf

(49 downloads)

Keywords

Subscribe

[1] , "End to end spoofing detection with raw wave CLDNNs", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1649. Accessed: Jun. 28, 2017.
@article{1649-17,
url = {http://sigport.org/1649},
author = { },
publisher = {IEEE SigPort},
title = {End to end spoofing detection with raw wave CLDNNs},
year = {2017} }
TY - EJOUR
T1 - End to end spoofing detection with raw wave CLDNNs
AU -
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1649
ER -
. (2017). End to end spoofing detection with raw wave CLDNNs. IEEE SigPort. http://sigport.org/1649
, 2017. End to end spoofing detection with raw wave CLDNNs. Available at: http://sigport.org/1649.
. (2017). "End to end spoofing detection with raw wave CLDNNs." Web.
1. . End to end spoofing detection with raw wave CLDNNs [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1649

Automatic Radar Waveform Recognition Based on Time-Frequency Analysis and Convolutional Neural Network

Paper Details

Authors:
Chao Wang, Jian Wang, Xudong Zhang
Submitted On:
3 March 2017 - 10:02am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Poster_1571.pdf

(44 downloads)

Keywords

Subscribe

[1] Chao Wang, Jian Wang, Xudong Zhang, "Automatic Radar Waveform Recognition Based on Time-Frequency Analysis and Convolutional Neural Network", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1614. Accessed: Jun. 28, 2017.
@article{1614-17,
url = {http://sigport.org/1614},
author = {Chao Wang; Jian Wang; Xudong Zhang },
publisher = {IEEE SigPort},
title = {Automatic Radar Waveform Recognition Based on Time-Frequency Analysis and Convolutional Neural Network},
year = {2017} }
TY - EJOUR
T1 - Automatic Radar Waveform Recognition Based on Time-Frequency Analysis and Convolutional Neural Network
AU - Chao Wang; Jian Wang; Xudong Zhang
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1614
ER -
Chao Wang, Jian Wang, Xudong Zhang. (2017). Automatic Radar Waveform Recognition Based on Time-Frequency Analysis and Convolutional Neural Network. IEEE SigPort. http://sigport.org/1614
Chao Wang, Jian Wang, Xudong Zhang, 2017. Automatic Radar Waveform Recognition Based on Time-Frequency Analysis and Convolutional Neural Network. Available at: http://sigport.org/1614.
Chao Wang, Jian Wang, Xudong Zhang. (2017). "Automatic Radar Waveform Recognition Based on Time-Frequency Analysis and Convolutional Neural Network." Web.
1. Chao Wang, Jian Wang, Xudong Zhang. Automatic Radar Waveform Recognition Based on Time-Frequency Analysis and Convolutional Neural Network [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1614

An Online Feature Selection Architecture For Human Activity Recognition


Human Activity Recognition (HAR) must currently face up to the challenge of rethinking analytics from the perspective of real-time operation, wherein biophysical sensing streams are efficiently intertwined at close vicinity to the point of sensing. As such, feature selection techniques, traditionally employed for off-line data processing, should be evaluated with respect to their ability to filter out redundant information in real-time.

Paper Details

Authors:
Athanasia Panousopoulou, Panagiotis Tsakalides
Submitted On:
3 March 2017 - 9:28am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

karagiannaki_ICASSP_2017_POSTER.pdf

(54 downloads)

Keywords

Subscribe

[1] Athanasia Panousopoulou, Panagiotis Tsakalides, "An Online Feature Selection Architecture For Human Activity Recognition", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1613. Accessed: Jun. 28, 2017.
@article{1613-17,
url = {http://sigport.org/1613},
author = {Athanasia Panousopoulou; Panagiotis Tsakalides },
publisher = {IEEE SigPort},
title = {An Online Feature Selection Architecture For Human Activity Recognition},
year = {2017} }
TY - EJOUR
T1 - An Online Feature Selection Architecture For Human Activity Recognition
AU - Athanasia Panousopoulou; Panagiotis Tsakalides
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1613
ER -
Athanasia Panousopoulou, Panagiotis Tsakalides. (2017). An Online Feature Selection Architecture For Human Activity Recognition. IEEE SigPort. http://sigport.org/1613
Athanasia Panousopoulou, Panagiotis Tsakalides, 2017. An Online Feature Selection Architecture For Human Activity Recognition. Available at: http://sigport.org/1613.
Athanasia Panousopoulou, Panagiotis Tsakalides. (2017). "An Online Feature Selection Architecture For Human Activity Recognition." Web.
1. Athanasia Panousopoulou, Panagiotis Tsakalides. An Online Feature Selection Architecture For Human Activity Recognition [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1613

Part-Level Fully Convolutional Networks for Pedestrian Detection


Since pedestrians in videos have a wide range of appearances such as body poses, occlusions, and complex backgrounds, pedestrian detection is a challengeable task. In this paper, we propose part-level fully convolutional networks (FCN) for pedestrian detection. We adopt deep learning to deal with the proposal shifting problem in pedestrian detection. First, we combine convolutional neural networks (CNN) and FCN to align bounding boxes for pedestrians. Then, we perform part-level pedestrian detection based on CNN to recall the lost body parts.

Paper Details

Authors:
Xinran Wang,Cheolkon Jung,Alfred O Hero
Submitted On:
8 March 2017 - 10:41pm
Short Link:
Type:
Event:
Document Year:
Cite

Document Files

keynote

(68 downloads)

Keywords

Subscribe

[1] Xinran Wang,Cheolkon Jung,Alfred O Hero, "Part-Level Fully Convolutional Networks for Pedestrian Detection", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1570. Accessed: Jun. 28, 2017.
@article{1570-17,
url = {http://sigport.org/1570},
author = {Xinran Wang;Cheolkon Jung;Alfred O Hero },
publisher = {IEEE SigPort},
title = {Part-Level Fully Convolutional Networks for Pedestrian Detection},
year = {2017} }
TY - EJOUR
T1 - Part-Level Fully Convolutional Networks for Pedestrian Detection
AU - Xinran Wang;Cheolkon Jung;Alfred O Hero
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1570
ER -
Xinran Wang,Cheolkon Jung,Alfred O Hero. (2017). Part-Level Fully Convolutional Networks for Pedestrian Detection. IEEE SigPort. http://sigport.org/1570
Xinran Wang,Cheolkon Jung,Alfred O Hero, 2017. Part-Level Fully Convolutional Networks for Pedestrian Detection. Available at: http://sigport.org/1570.
Xinran Wang,Cheolkon Jung,Alfred O Hero. (2017). "Part-Level Fully Convolutional Networks for Pedestrian Detection." Web.
1. Xinran Wang,Cheolkon Jung,Alfred O Hero. Part-Level Fully Convolutional Networks for Pedestrian Detection [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1570

TWO-STAGE FACIAL AGE PREDICTION USING GROUP-SPECIFIC FEATURES


A novel two-stage age prediction approach with group-specific features is proposed in this paper. Aging process is captured through a highly discriminating feature representation that models shape, appearance, skin spots, and wrinkles. The two-stage method consists of a multi-class Support Vector Machine (SVM) to predict the age bracket while the final age prediction is carried out using Support Vector Regression (SVR). The novelty of our work is that the feature extraction is group-specific and can therefore be tailored to each age bracket in the specific age prediction step.

Paper Details

Authors:
Jhony K. Pontes, Clinton Fookes, Alceu S. Britto Jr., Alessandro L. Koerich
Submitted On:
1 March 2017 - 6:07pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICASSP2017.pdf

(49 downloads)

Keywords

Subscribe

[1] Jhony K. Pontes, Clinton Fookes, Alceu S. Britto Jr., Alessandro L. Koerich, "TWO-STAGE FACIAL AGE PREDICTION USING GROUP-SPECIFIC FEATURES", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1564. Accessed: Jun. 28, 2017.
@article{1564-17,
url = {http://sigport.org/1564},
author = {Jhony K. Pontes; Clinton Fookes; Alceu S. Britto Jr.; Alessandro L. Koerich },
publisher = {IEEE SigPort},
title = {TWO-STAGE FACIAL AGE PREDICTION USING GROUP-SPECIFIC FEATURES},
year = {2017} }
TY - EJOUR
T1 - TWO-STAGE FACIAL AGE PREDICTION USING GROUP-SPECIFIC FEATURES
AU - Jhony K. Pontes; Clinton Fookes; Alceu S. Britto Jr.; Alessandro L. Koerich
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1564
ER -
Jhony K. Pontes, Clinton Fookes, Alceu S. Britto Jr., Alessandro L. Koerich. (2017). TWO-STAGE FACIAL AGE PREDICTION USING GROUP-SPECIFIC FEATURES. IEEE SigPort. http://sigport.org/1564
Jhony K. Pontes, Clinton Fookes, Alceu S. Britto Jr., Alessandro L. Koerich, 2017. TWO-STAGE FACIAL AGE PREDICTION USING GROUP-SPECIFIC FEATURES. Available at: http://sigport.org/1564.
Jhony K. Pontes, Clinton Fookes, Alceu S. Britto Jr., Alessandro L. Koerich. (2017). "TWO-STAGE FACIAL AGE PREDICTION USING GROUP-SPECIFIC FEATURES." Web.
1. Jhony K. Pontes, Clinton Fookes, Alceu S. Britto Jr., Alessandro L. Koerich. TWO-STAGE FACIAL AGE PREDICTION USING GROUP-SPECIFIC FEATURES [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1564

Through-the-Wall Radar Signal Classification using Discriminative Dictionary Learning


Through-the-wall radar imaging is an electromagnetic wave sensing technology capable of detecting targets behind walls, doors, and opaque obstacles. Identification of stationary targets is often achieved by first forming an image of the scene, and then segmenting and classifying the targets of interest. In order to provide prompt and reliable situational awareness, this paper proposes a radar signal classification approach that does not rely on image formation.

Paper Details

Authors:
Abdesselam Bouzerdoum, Fok Hing Chi Tivive, and Jia Fei
Submitted On:
1 March 2017 - 4:33pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

ICASSP2017 v2.pptx

(38 downloads)

Keywords

Subscribe

[1] Abdesselam Bouzerdoum, Fok Hing Chi Tivive, and Jia Fei, "Through-the-Wall Radar Signal Classification using Discriminative Dictionary Learning", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1563. Accessed: Jun. 28, 2017.
@article{1563-17,
url = {http://sigport.org/1563},
author = {Abdesselam Bouzerdoum; Fok Hing Chi Tivive; and Jia Fei },
publisher = {IEEE SigPort},
title = {Through-the-Wall Radar Signal Classification using Discriminative Dictionary Learning},
year = {2017} }
TY - EJOUR
T1 - Through-the-Wall Radar Signal Classification using Discriminative Dictionary Learning
AU - Abdesselam Bouzerdoum; Fok Hing Chi Tivive; and Jia Fei
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1563
ER -
Abdesselam Bouzerdoum, Fok Hing Chi Tivive, and Jia Fei. (2017). Through-the-Wall Radar Signal Classification using Discriminative Dictionary Learning. IEEE SigPort. http://sigport.org/1563
Abdesselam Bouzerdoum, Fok Hing Chi Tivive, and Jia Fei, 2017. Through-the-Wall Radar Signal Classification using Discriminative Dictionary Learning. Available at: http://sigport.org/1563.
Abdesselam Bouzerdoum, Fok Hing Chi Tivive, and Jia Fei. (2017). "Through-the-Wall Radar Signal Classification using Discriminative Dictionary Learning." Web.
1. Abdesselam Bouzerdoum, Fok Hing Chi Tivive, and Jia Fei. Through-the-Wall Radar Signal Classification using Discriminative Dictionary Learning [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1563

FAST SPECTRAL CLUSTERING WITH EFFICIENT LARGE GRAPH CONSTRUCTION

Paper Details

Authors:
Wei Zhu, Feiping Nie, Xuelong Li
Submitted On:
1 March 2017 - 6:05am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

poster of FSC

(43 downloads)

Keywords

Subscribe

[1] Wei Zhu, Feiping Nie, Xuelong Li, " FAST SPECTRAL CLUSTERING WITH EFFICIENT LARGE GRAPH CONSTRUCTION", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1522. Accessed: Jun. 28, 2017.
@article{1522-17,
url = {http://sigport.org/1522},
author = {Wei Zhu; Feiping Nie; Xuelong Li },
publisher = {IEEE SigPort},
title = { FAST SPECTRAL CLUSTERING WITH EFFICIENT LARGE GRAPH CONSTRUCTION},
year = {2017} }
TY - EJOUR
T1 - FAST SPECTRAL CLUSTERING WITH EFFICIENT LARGE GRAPH CONSTRUCTION
AU - Wei Zhu; Feiping Nie; Xuelong Li
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1522
ER -
Wei Zhu, Feiping Nie, Xuelong Li. (2017). FAST SPECTRAL CLUSTERING WITH EFFICIENT LARGE GRAPH CONSTRUCTION. IEEE SigPort. http://sigport.org/1522
Wei Zhu, Feiping Nie, Xuelong Li, 2017. FAST SPECTRAL CLUSTERING WITH EFFICIENT LARGE GRAPH CONSTRUCTION. Available at: http://sigport.org/1522.
Wei Zhu, Feiping Nie, Xuelong Li. (2017). " FAST SPECTRAL CLUSTERING WITH EFFICIENT LARGE GRAPH CONSTRUCTION." Web.
1. Wei Zhu, Feiping Nie, Xuelong Li. FAST SPECTRAL CLUSTERING WITH EFFICIENT LARGE GRAPH CONSTRUCTION [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1522

Real-time Audio Classifi cation based on Mixture Models

Paper Details

Authors:
Christophe Biernacki, Raphaël Greff
Submitted On:
28 February 2017 - 8:33am
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Poster of Maxime Baelde for ICASSP 2017

(41 downloads)

Keywords

Subscribe

[1] Christophe Biernacki, Raphaël Greff, "Real-time Audio Classifi cation based on Mixture Models", IEEE SigPort, 2017. [Online]. Available: http://sigport.org/1510. Accessed: Jun. 28, 2017.
@article{1510-17,
url = {http://sigport.org/1510},
author = {Christophe Biernacki; Raphaël Greff },
publisher = {IEEE SigPort},
title = {Real-time Audio Classifi cation based on Mixture Models},
year = {2017} }
TY - EJOUR
T1 - Real-time Audio Classifi cation based on Mixture Models
AU - Christophe Biernacki; Raphaël Greff
PY - 2017
PB - IEEE SigPort
UR - http://sigport.org/1510
ER -
Christophe Biernacki, Raphaël Greff. (2017). Real-time Audio Classifi cation based on Mixture Models. IEEE SigPort. http://sigport.org/1510
Christophe Biernacki, Raphaël Greff, 2017. Real-time Audio Classifi cation based on Mixture Models. Available at: http://sigport.org/1510.
Christophe Biernacki, Raphaël Greff. (2017). "Real-time Audio Classifi cation based on Mixture Models." Web.
1. Christophe Biernacki, Raphaël Greff. Real-time Audio Classifi cation based on Mixture Models [Internet]. IEEE SigPort; 2017. Available from : http://sigport.org/1510

A ROBUST APPLICATION DETECTOR FOR INTELLIGENT WIRELESS COLLABORATION


We present an approach for detecting application level protocols over a wireless communications link, without the need for demodulation or decryption. Our detector is suitable for diverse radio types, since only simple external signal features are used as inputs. We show that the Profile Hidden Markov Model (PHMM) is well suited to this task, due to the probabilistic nature of the wireless channel and the discrete nature of application level traffic. We include results evaluating the detection performance for two application protocols in 802.11 in the presence of background traffic.

Paper Details

Authors:
Kevin Pietsch, Sean Mason
Submitted On:
8 December 2016 - 1:03pm
Short Link:
Type:
Event:
Presenter's Name:
Paper Code:
Document Year:
Cite

Document Files

Presentation Slides

(60 downloads)

Paper

(72 downloads)

Keywords

Subscribe

[1] Kevin Pietsch, Sean Mason, "A ROBUST APPLICATION DETECTOR FOR INTELLIGENT WIRELESS COLLABORATION", IEEE SigPort, 2016. [Online]. Available: http://sigport.org/1422. Accessed: Jun. 28, 2017.
@article{1422-16,
url = {http://sigport.org/1422},
author = {Kevin Pietsch; Sean Mason },
publisher = {IEEE SigPort},
title = {A ROBUST APPLICATION DETECTOR FOR INTELLIGENT WIRELESS COLLABORATION},
year = {2016} }
TY - EJOUR
T1 - A ROBUST APPLICATION DETECTOR FOR INTELLIGENT WIRELESS COLLABORATION
AU - Kevin Pietsch; Sean Mason
PY - 2016
PB - IEEE SigPort
UR - http://sigport.org/1422
ER -
Kevin Pietsch, Sean Mason. (2016). A ROBUST APPLICATION DETECTOR FOR INTELLIGENT WIRELESS COLLABORATION. IEEE SigPort. http://sigport.org/1422
Kevin Pietsch, Sean Mason, 2016. A ROBUST APPLICATION DETECTOR FOR INTELLIGENT WIRELESS COLLABORATION. Available at: http://sigport.org/1422.
Kevin Pietsch, Sean Mason. (2016). "A ROBUST APPLICATION DETECTOR FOR INTELLIGENT WIRELESS COLLABORATION." Web.
1. Kevin Pietsch, Sean Mason. A ROBUST APPLICATION DETECTOR FOR INTELLIGENT WIRELESS COLLABORATION [Internet]. IEEE SigPort; 2016. Available from : http://sigport.org/1422

Pages