Target detection using incremental learning on single-trial evoked response

Yonghong Huang, Deniz Erdogmus, Misha Pavel, Kenneth E. Hild, Santosh Mathan

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The human neural responses associated with cognitive events, referred as event related potentials (ERPs), can provide reliable inference for target image detection. Incremental learning has been widely investigated to deal with large datasets. To solve the problem of data growing over time in cross session studies, we apply an incremental learning support vector machines (SVM) method on single-trial ERP detection for identifying targets in satellite images. We implement the incremental learning SVM by keeping only the support vectors, instead of all the data, from the previous sessions and incorporating them with the data of the current session. Thus the incremental learning dramatically reduces the computational load. The results demonstrate that the incremental learning ERP detection system performs as well as the naive method, which uses only the current training session, and the batch mode, which uses all training data. Furthermore, it is more computationally efficient, which allows it to better cope with a continuous stream of EEG data.

Original languageEnglish (US)
Title of host publicationICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Pages481-484
Number of pages4
StatePublished - 2009
Event2009 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2009 - Taipei, Taiwan, Province of China
Duration: Apr 19 2009Apr 24 2009

Other

Other2009 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2009
CountryTaiwan, Province of China
CityTaipei
Period4/19/094/24/09

Fingerprint

Target tracking
Support vector machines
Electroencephalography
Satellites

Keywords

  • Brain computer interface
  • Event-related potential
  • Incremental learning
  • Support vector machine
  • Target detection

ASJC Scopus subject areas

  • Signal Processing
  • Software
  • Electrical and Electronic Engineering

Cite this

Huang, Y., Erdogmus, D., Pavel, M., Hild, K. E., & Mathan, S. (2009). Target detection using incremental learning on single-trial evoked response. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings (pp. 481-484). [4959625]

Target detection using incremental learning on single-trial evoked response. / Huang, Yonghong; Erdogmus, Deniz; Pavel, Misha; Hild, Kenneth E.; Mathan, Santosh.

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2009. p. 481-484 4959625.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Huang, Y, Erdogmus, D, Pavel, M, Hild, KE & Mathan, S 2009, Target detection using incremental learning on single-trial evoked response. in ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings., 4959625, pp. 481-484, 2009 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2009, Taipei, Taiwan, Province of China, 4/19/09.
Huang Y, Erdogmus D, Pavel M, Hild KE, Mathan S. Target detection using incremental learning on single-trial evoked response. In ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2009. p. 481-484. 4959625
Huang, Yonghong ; Erdogmus, Deniz ; Pavel, Misha ; Hild, Kenneth E. ; Mathan, Santosh. / Target detection using incremental learning on single-trial evoked response. ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. 2009. pp. 481-484
@inproceedings{0f820e739cbd4728a61778eff2877a7e,
title = "Target detection using incremental learning on single-trial evoked response",
abstract = "The human neural responses associated with cognitive events, referred as event related potentials (ERPs), can provide reliable inference for target image detection. Incremental learning has been widely investigated to deal with large datasets. To solve the problem of data growing over time in cross session studies, we apply an incremental learning support vector machines (SVM) method on single-trial ERP detection for identifying targets in satellite images. We implement the incremental learning SVM by keeping only the support vectors, instead of all the data, from the previous sessions and incorporating them with the data of the current session. Thus the incremental learning dramatically reduces the computational load. The results demonstrate that the incremental learning ERP detection system performs as well as the naive method, which uses only the current training session, and the batch mode, which uses all training data. Furthermore, it is more computationally efficient, which allows it to better cope with a continuous stream of EEG data.",
keywords = "Brain computer interface, Event-related potential, Incremental learning, Support vector machine, Target detection",
author = "Yonghong Huang and Deniz Erdogmus and Misha Pavel and Hild, {Kenneth E.} and Santosh Mathan",
year = "2009",
language = "English (US)",
isbn = "9781424423545",
pages = "481--484",
booktitle = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

}

TY - GEN

T1 - Target detection using incremental learning on single-trial evoked response

AU - Huang, Yonghong

AU - Erdogmus, Deniz

AU - Pavel, Misha

AU - Hild, Kenneth E.

AU - Mathan, Santosh

PY - 2009

Y1 - 2009

N2 - The human neural responses associated with cognitive events, referred as event related potentials (ERPs), can provide reliable inference for target image detection. Incremental learning has been widely investigated to deal with large datasets. To solve the problem of data growing over time in cross session studies, we apply an incremental learning support vector machines (SVM) method on single-trial ERP detection for identifying targets in satellite images. We implement the incremental learning SVM by keeping only the support vectors, instead of all the data, from the previous sessions and incorporating them with the data of the current session. Thus the incremental learning dramatically reduces the computational load. The results demonstrate that the incremental learning ERP detection system performs as well as the naive method, which uses only the current training session, and the batch mode, which uses all training data. Furthermore, it is more computationally efficient, which allows it to better cope with a continuous stream of EEG data.

AB - The human neural responses associated with cognitive events, referred as event related potentials (ERPs), can provide reliable inference for target image detection. Incremental learning has been widely investigated to deal with large datasets. To solve the problem of data growing over time in cross session studies, we apply an incremental learning support vector machines (SVM) method on single-trial ERP detection for identifying targets in satellite images. We implement the incremental learning SVM by keeping only the support vectors, instead of all the data, from the previous sessions and incorporating them with the data of the current session. Thus the incremental learning dramatically reduces the computational load. The results demonstrate that the incremental learning ERP detection system performs as well as the naive method, which uses only the current training session, and the batch mode, which uses all training data. Furthermore, it is more computationally efficient, which allows it to better cope with a continuous stream of EEG data.

KW - Brain computer interface

KW - Event-related potential

KW - Incremental learning

KW - Support vector machine

KW - Target detection

UR - http://www.scopus.com/inward/record.url?scp=70349454107&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70349454107&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9781424423545

SP - 481

EP - 484

BT - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

ER -