Deep neural networks for kitchen activity recognition

Juarez Monteiro; Roger Granada; Rodrigo C Barros; Felipe Meneguzzi

doi:10.1109/IJCNN.2017.7966102

Deep neural networks for kitchen activity recognition

Juarez Monteiro, Roger Granada, Rodrigo C Barros, Felipe Meneguzzi

Computing Science

Pontifícia Universidade Católica do Rio Grande do Sul

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

23 Citations (Scopus)

Abstract

With the growth of video content produced by mobile cameras and surveillance systems, an increasing amount of data is becoming available and can be used for a variety of applications such as video surveillance, smart homes, smart cities, and in-home elder monitoring. Such applications focus in recognizing human activities in order to perform different tasks allowing the opportunity to support people in their different scenarios. In this paper we propose a deep neural architecture for kitchen human action recognition. This architecture contains an ensemble of convolutional neural networks connected through different fusion methods to predict the label of each action. Experiments show that our architecture achieves the novel state-of-the-art for identifying cooking actions in a well-known kitchen dataset.

Original language	English
Title of host publication	Neural Networks (IJCNN), 2017 International Joint Conference on
Publisher	IEEE Explore
Pages	2048-2055
Number of pages	8
ISBN (Electronic)	978-1-5090-6182-2
ISBN (Print)	978-1-5090-6183-9
DOIs	https://doi.org/10.1109/IJCNN.2017.7966102
Publication status	Published - 3 Jul 2017
Event	2017 International Joint Conference on Neural Networks (IJCNN) - Anchorage, AK, USA, Anchorage, United States Duration: 14 May 2017 → 17 May 2017

Conference

Conference	2017 International Joint Conference on Neural Networks (IJCNN)
Country/Territory	United States
City	Anchorage
Period	14/05/17 → 17/05/17

Bibliographical note

ACKNOWLEDGEMENT
This paper was achieved in cooperation with HP Brasil Indústria e Comércio de Equipamentos Eletrônicos LTDA. using incentives of Brazilian Informatics Law (Law № 8.2.48 of 1991). The authors also would like to thank FAPERGS, CNPq, and CAPES for funding this research.

Keywords

neural networks
Computer architecture
Support vector machines
optical imaging
activity recognition
cameras

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/IJCNN.2017.7966102Licence: Unspecified

Cite this

@inproceedings{db81dee1b28843128042e8e39d3362e1,

title = "Deep neural networks for kitchen activity recognition",

abstract = "With the growth of video content produced by mobile cameras and surveillance systems, an increasing amount of data is becoming available and can be used for a variety of applications such as video surveillance, smart homes, smart cities, and in-home elder monitoring. Such applications focus in recognizing human activities in order to perform different tasks allowing the opportunity to support people in their different scenarios. In this paper we propose a deep neural architecture for kitchen human action recognition. This architecture contains an ensemble of convolutional neural networks connected through different fusion methods to predict the label of each action. Experiments show that our architecture achieves the novel state-of-the-art for identifying cooking actions in a well-known kitchen dataset.",

keywords = "neural networks, Computer architecture, Support vector machines, optical imaging, activity recognition, cameras",

author = "Juarez Monteiro and Roger Granada and Barros, {Rodrigo C} and Felipe Meneguzzi",

note = "ACKNOWLEDGEMENT This paper was achieved in cooperation with HP Brasil Ind{\'u}stria e Com{\'e}rcio de Equipamentos Eletr{\^o}nicos LTDA. using incentives of Brazilian Informatics Law (Law № 8.2.48 of 1991). The authors also would like to thank FAPERGS, CNPq, and CAPES for funding this research.; 2017 International Joint Conference on Neural Networks (IJCNN) ; Conference date: 14-05-2017 Through 17-05-2017",

year = "2017",

month = jul,

day = "3",

doi = "10.1109/IJCNN.2017.7966102",

language = "English",

isbn = "978-1-5090-6183-9",

pages = "2048--2055",

booktitle = "Neural Networks (IJCNN), 2017 International Joint Conference on",

publisher = "IEEE Explore",

}

TY - GEN

T1 - Deep neural networks for kitchen activity recognition

AU - Monteiro, Juarez

AU - Granada, Roger

AU - Barros, Rodrigo C

AU - Meneguzzi, Felipe

N1 - ACKNOWLEDGEMENT This paper was achieved in cooperation with HP Brasil Indústria e Comércio de Equipamentos Eletrônicos LTDA. using incentives of Brazilian Informatics Law (Law № 8.2.48 of 1991). The authors also would like to thank FAPERGS, CNPq, and CAPES for funding this research.

PY - 2017/7/3

Y1 - 2017/7/3

N2 - With the growth of video content produced by mobile cameras and surveillance systems, an increasing amount of data is becoming available and can be used for a variety of applications such as video surveillance, smart homes, smart cities, and in-home elder monitoring. Such applications focus in recognizing human activities in order to perform different tasks allowing the opportunity to support people in their different scenarios. In this paper we propose a deep neural architecture for kitchen human action recognition. This architecture contains an ensemble of convolutional neural networks connected through different fusion methods to predict the label of each action. Experiments show that our architecture achieves the novel state-of-the-art for identifying cooking actions in a well-known kitchen dataset.

AB - With the growth of video content produced by mobile cameras and surveillance systems, an increasing amount of data is becoming available and can be used for a variety of applications such as video surveillance, smart homes, smart cities, and in-home elder monitoring. Such applications focus in recognizing human activities in order to perform different tasks allowing the opportunity to support people in their different scenarios. In this paper we propose a deep neural architecture for kitchen human action recognition. This architecture contains an ensemble of convolutional neural networks connected through different fusion methods to predict the label of each action. Experiments show that our architecture achieves the novel state-of-the-art for identifying cooking actions in a well-known kitchen dataset.

KW - neural networks

KW - Computer architecture

KW - Support vector machines

KW - optical imaging

KW - activity recognition

KW - cameras

U2 - 10.1109/IJCNN.2017.7966102

DO - 10.1109/IJCNN.2017.7966102

M3 - Published conference contribution

SN - 978-1-5090-6183-9

SP - 2048

EP - 2055

BT - Neural Networks (IJCNN), 2017 International Joint Conference on

PB - IEEE Explore

T2 - 2017 International Joint Conference on Neural Networks (IJCNN)

Y2 - 14 May 2017 through 17 May 2017

ER -

Deep neural networks for kitchen activity recognition

Abstract

Conference

Bibliographical note

Keywords

UN SDGs

Access to Document

Fingerprint

Cite this