Factored four way conditional restricted Boltzmann machines for activity recognition

Decebal Constantin Mocanu, Haitham Bou Ammar, Dietwig Lowet, Kurt Driessens, Antonio Liotta, Gerhard Weiss, Karl Tuyls

Research output: Contribution to journalArticleAcademicpeer-review

9 Citations (Scopus)

Abstract

This paper introduces a new learning algorithm for human activity recognition capable of simultaneous regression and classification. Building upon conditional restricted Boltzmann machines (CRBMs), Factored four way conditional restricted Boltzmann machines (FFW-CRBMs) incorporate a new label layer and four-way interactions among the neurons from the different layers. The additional layer gives the classification nodes a similar strong multiplicative effect compared to the other layers, and avoids that the classification neurons are overwhelmed by the (much larger set of) other neurons. This makes FFW-CRBMs capable of performing activity recognition, prediction and self auto evaluation of classification within one unified framework. As a second contribution, sequential Markov chain contrastive divergence (SMcCD) is introduced. SMcCD modifies Contrastive Divergence to compensate for the extra complexity of FFW-CRBMs during training. Two sets of experiments one on benchmark datasets and one a robotic platform for smart companions show the effectiveness of FFW-CRBMs.
Original languageEnglish
Pages (from-to)100-108
Number of pages9
JournalPattern Recognition Letters
Volume66
DOIs
Publication statusPublished - 15 Nov 2015

Keywords

  • Activity recognition
  • Deep learning
  • Restricted Boltzmann machines

Cite this

Mocanu, Decebal Constantin ; Bou Ammar, Haitham ; Lowet, Dietwig ; Driessens, Kurt ; Liotta, Antonio ; Weiss, Gerhard ; Tuyls, Karl. / Factored four way conditional restricted Boltzmann machines for activity recognition. In: Pattern Recognition Letters. 2015 ; Vol. 66. pp. 100-108.
@article{b0de81e7d4d94e3dbe381bc647bb5bc6,
title = "Factored four way conditional restricted Boltzmann machines for activity recognition",
abstract = "This paper introduces a new learning algorithm for human activity recognition capable of simultaneous regression and classification. Building upon conditional restricted Boltzmann machines (CRBMs), Factored four way conditional restricted Boltzmann machines (FFW-CRBMs) incorporate a new label layer and four-way interactions among the neurons from the different layers. The additional layer gives the classification nodes a similar strong multiplicative effect compared to the other layers, and avoids that the classification neurons are overwhelmed by the (much larger set of) other neurons. This makes FFW-CRBMs capable of performing activity recognition, prediction and self auto evaluation of classification within one unified framework. As a second contribution, sequential Markov chain contrastive divergence (SMcCD) is introduced. SMcCD modifies Contrastive Divergence to compensate for the extra complexity of FFW-CRBMs during training. Two sets of experiments one on benchmark datasets and one a robotic platform for smart companions show the effectiveness of FFW-CRBMs.",
keywords = "Activity recognition, Deep learning, Restricted Boltzmann machines",
author = "Mocanu, {Decebal Constantin} and {Bou Ammar}, Haitham and Dietwig Lowet and Kurt Driessens and Antonio Liotta and Gerhard Weiss and Karl Tuyls",
year = "2015",
month = "11",
day = "15",
doi = "10.1016/j.patrec.2015.01.013",
language = "English",
volume = "66",
pages = "100--108",
journal = "Pattern Recognition Letters",
issn = "0167-8655",
publisher = "Elsevier",

}

Factored four way conditional restricted Boltzmann machines for activity recognition. / Mocanu, Decebal Constantin; Bou Ammar, Haitham; Lowet, Dietwig; Driessens, Kurt; Liotta, Antonio; Weiss, Gerhard; Tuyls, Karl.

In: Pattern Recognition Letters, Vol. 66, 15.11.2015, p. 100-108.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Factored four way conditional restricted Boltzmann machines for activity recognition

AU - Mocanu, Decebal Constantin

AU - Bou Ammar, Haitham

AU - Lowet, Dietwig

AU - Driessens, Kurt

AU - Liotta, Antonio

AU - Weiss, Gerhard

AU - Tuyls, Karl

PY - 2015/11/15

Y1 - 2015/11/15

N2 - This paper introduces a new learning algorithm for human activity recognition capable of simultaneous regression and classification. Building upon conditional restricted Boltzmann machines (CRBMs), Factored four way conditional restricted Boltzmann machines (FFW-CRBMs) incorporate a new label layer and four-way interactions among the neurons from the different layers. The additional layer gives the classification nodes a similar strong multiplicative effect compared to the other layers, and avoids that the classification neurons are overwhelmed by the (much larger set of) other neurons. This makes FFW-CRBMs capable of performing activity recognition, prediction and self auto evaluation of classification within one unified framework. As a second contribution, sequential Markov chain contrastive divergence (SMcCD) is introduced. SMcCD modifies Contrastive Divergence to compensate for the extra complexity of FFW-CRBMs during training. Two sets of experiments one on benchmark datasets and one a robotic platform for smart companions show the effectiveness of FFW-CRBMs.

AB - This paper introduces a new learning algorithm for human activity recognition capable of simultaneous regression and classification. Building upon conditional restricted Boltzmann machines (CRBMs), Factored four way conditional restricted Boltzmann machines (FFW-CRBMs) incorporate a new label layer and four-way interactions among the neurons from the different layers. The additional layer gives the classification nodes a similar strong multiplicative effect compared to the other layers, and avoids that the classification neurons are overwhelmed by the (much larger set of) other neurons. This makes FFW-CRBMs capable of performing activity recognition, prediction and self auto evaluation of classification within one unified framework. As a second contribution, sequential Markov chain contrastive divergence (SMcCD) is introduced. SMcCD modifies Contrastive Divergence to compensate for the extra complexity of FFW-CRBMs during training. Two sets of experiments one on benchmark datasets and one a robotic platform for smart companions show the effectiveness of FFW-CRBMs.

KW - Activity recognition

KW - Deep learning

KW - Restricted Boltzmann machines

U2 - 10.1016/j.patrec.2015.01.013

DO - 10.1016/j.patrec.2015.01.013

M3 - Article

VL - 66

SP - 100

EP - 108

JO - Pattern Recognition Letters

JF - Pattern Recognition Letters

SN - 0167-8655

ER -