Boosted convolutional neural networks

Publikation: Konferencebidrag › Paper › Forskning › fagfællebedømt

Standard

Boosted convolutional neural networks. / Moghimi, Mohammad; Saberian, Mohammad; Yang, Jian; Li, Li Jia; Vasconcelos, Nuno; Belongie, Serge.

2016. 24.1-24.13 Paper præsenteret ved 27th British Machine Vision Conference, BMVC 2016, York, Storbritannien.

Publikation: Konferencebidrag › Paper › Forskning › fagfællebedømt

Harvard

Moghimi, M, Saberian, M, Yang, J, Li, LJ, Vasconcelos, N & Belongie, S 2016, 'Boosted convolutional neural networks', Paper fremlagt ved 27th British Machine Vision Conference, BMVC 2016, York, Storbritannien, 19/09/2016 - 22/09/2016 s. 24.1-24.13.

APA

Moghimi, M., Saberian, M., Yang, J., Li, L. J., Vasconcelos, N., & Belongie, S. (2016). Boosted convolutional neural networks. 24.1-24.13. Paper præsenteret ved 27th British Machine Vision Conference, BMVC 2016, York, Storbritannien.

Vancouver

Moghimi M, Saberian M, Yang J, Li LJ, Vasconcelos N, Belongie S. Boosted convolutional neural networks. 2016. Paper præsenteret ved 27th British Machine Vision Conference, BMVC 2016, York, Storbritannien.

Author

Moghimi, Mohammad ; Saberian, Mohammad ; Yang, Jian ; Li, Li Jia ; Vasconcelos, Nuno ; Belongie, Serge. / Boosted convolutional neural networks. Paper præsenteret ved 27th British Machine Vision Conference, BMVC 2016, York, Storbritannien.

Bibtex

@conference{cdb6b625b76c48bf86379e536dcba182,

title = "Boosted convolutional neural networks",

abstract = "In this work, we propose a new algorithm for boosting Deep Convolutional Neural Networks (BoostCNN) to combine the merits of boosting and these networks. To learn this new model, we propose a novel algorithm to incorporate boosting weights into the deep learning architecture based on least square objective function. We also show that it is possible to use networks of different structures within the proposed boosting framework and BoostCNN is able to select the best network structure in each iteration. This not only results in superior performance but also reduces the required manual effort for finding the right network structure. Experiments show that the proposed method is able to achieve state-of-the-art performance on several fine-grained classification tasks such as bird, car, and aircraft classification.",

author = "Mohammad Moghimi and Mohammad Saberian and Jian Yang and Li, {Li Jia} and Nuno Vasconcelos and Serge Belongie",

note = "Publisher Copyright: {\textcopyright} 2016. The copyright of this document resides with its authors.; 27th British Machine Vision Conference, BMVC 2016 ; Conference date: 19-09-2016 Through 22-09-2016",

year = "2016",

language = "English",

pages = "24.1--24.13",

}

RIS

TY - CONF

T1 - Boosted convolutional neural networks

AU - Moghimi, Mohammad

AU - Saberian, Mohammad

AU - Yang, Jian

AU - Li, Li Jia

AU - Vasconcelos, Nuno

AU - Belongie, Serge

PY - 2016

Y1 - 2016

N2 - In this work, we propose a new algorithm for boosting Deep Convolutional Neural Networks (BoostCNN) to combine the merits of boosting and these networks. To learn this new model, we propose a novel algorithm to incorporate boosting weights into the deep learning architecture based on least square objective function. We also show that it is possible to use networks of different structures within the proposed boosting framework and BoostCNN is able to select the best network structure in each iteration. This not only results in superior performance but also reduces the required manual effort for finding the right network structure. Experiments show that the proposed method is able to achieve state-of-the-art performance on several fine-grained classification tasks such as bird, car, and aircraft classification.

AB - In this work, we propose a new algorithm for boosting Deep Convolutional Neural Networks (BoostCNN) to combine the merits of boosting and these networks. To learn this new model, we propose a novel algorithm to incorporate boosting weights into the deep learning architecture based on least square objective function. We also show that it is possible to use networks of different structures within the proposed boosting framework and BoostCNN is able to select the best network structure in each iteration. This not only results in superior performance but also reduces the required manual effort for finding the right network structure. Experiments show that the proposed method is able to achieve state-of-the-art performance on several fine-grained classification tasks such as bird, car, and aircraft classification.

UR - http://www.scopus.com/inward/record.url?scp=85047160033&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85047160033

SP - 24.1-24.13

T2 - 27th British Machine Vision Conference, BMVC 2016

Y2 - 19 September 2016 through 22 September 2016

ER -

ID: 301827868

Datalogisk Institut