Tampere University of Technology

TUTCRIS Research Portal

Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Standard

Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision. / Bai, Yue; Bhattacharyya, Shuvra S.; Happonen, Antti P.; Huttunen, Heikki.

2018 26th European Signal Processing Conference (EUSIPCO). IEEE, 2018.

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Harvard

Bai, Y, Bhattacharyya, SS, Happonen, AP & Huttunen, H 2018, Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision. in 2018 26th European Signal Processing Conference (EUSIPCO). IEEE, European Signal Processing Conference, 1/01/00. https://doi.org/10.23919/EUSIPCO.2018.8553186

APA

Bai, Y., Bhattacharyya, S. S., Happonen, A. P., & Huttunen, H. (2018). Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision. In 2018 26th European Signal Processing Conference (EUSIPCO) IEEE. https://doi.org/10.23919/EUSIPCO.2018.8553186

Vancouver

Bai Y, Bhattacharyya SS, Happonen AP, Huttunen H. Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision. In 2018 26th European Signal Processing Conference (EUSIPCO). IEEE. 2018 https://doi.org/10.23919/EUSIPCO.2018.8553186

Author

Bai, Yue ; Bhattacharyya, Shuvra S. ; Happonen, Antti P. ; Huttunen, Heikki. / Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision. 2018 26th European Signal Processing Conference (EUSIPCO). IEEE, 2018.

Bibtex - Download

@inproceedings{a2f6abb50acd416bbd4c47b9dde54c37,
title = "Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision",
abstract = "We propose a new framework for image classification with deep neural networks. The framework introduces intermediate outputs to the computational graph of a network. This enables flexible control of the computational load and balances the tradeoff between accuracy and execution time. Moreover, we present an interesting finding that the intermediate outputs can act as a regularizer at training time, improving the prediction accuracy. In the experimental section we demonstrate the performance of our proposed framework with various commonly used pretrained deep networks in the use case of apparent age estimation.",
author = "Yue Bai and Bhattacharyya, {Shuvra S.} and Happonen, {Antti P.} and Heikki Huttunen",
note = "INT=tie,{"}Bai, Yue{"}",
year = "2018",
month = "9",
doi = "10.23919/EUSIPCO.2018.8553186",
language = "English",
isbn = "978-1-5386-3736-4",
publisher = "IEEE",
booktitle = "2018 26th European Signal Processing Conference (EUSIPCO)",

}

RIS (suitable for import to EndNote) - Download

TY - GEN

T1 - Elastic Neural Networks: A Scalable Framework for Embedded Computer Vision

AU - Bai, Yue

AU - Bhattacharyya, Shuvra S.

AU - Happonen, Antti P.

AU - Huttunen, Heikki

N1 - INT=tie,"Bai, Yue"

PY - 2018/9

Y1 - 2018/9

N2 - We propose a new framework for image classification with deep neural networks. The framework introduces intermediate outputs to the computational graph of a network. This enables flexible control of the computational load and balances the tradeoff between accuracy and execution time. Moreover, we present an interesting finding that the intermediate outputs can act as a regularizer at training time, improving the prediction accuracy. In the experimental section we demonstrate the performance of our proposed framework with various commonly used pretrained deep networks in the use case of apparent age estimation.

AB - We propose a new framework for image classification with deep neural networks. The framework introduces intermediate outputs to the computational graph of a network. This enables flexible control of the computational load and balances the tradeoff between accuracy and execution time. Moreover, we present an interesting finding that the intermediate outputs can act as a regularizer at training time, improving the prediction accuracy. In the experimental section we demonstrate the performance of our proposed framework with various commonly used pretrained deep networks in the use case of apparent age estimation.

U2 - 10.23919/EUSIPCO.2018.8553186

DO - 10.23919/EUSIPCO.2018.8553186

M3 - Conference contribution

SN - 978-1-5386-3736-4

BT - 2018 26th European Signal Processing Conference (EUSIPCO)

PB - IEEE

ER -