Presentation | 2014-06-25 Implementation of Deep Neural Network on Image Classification Rui ZHONG, Taro TEZUKA, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In this research we implemented a scalable deep convolutional neural network based on GPU accelerating techniques, optimized the network architecture on GPU architecutre and reduced the training time by organizing device memory to taking advantages of parallel memory access. We adopted back propagation with limited kernel functions to get higher efficiency. Also we experiment on how learning rate parameters effect the deep network. Experiments on all cases of MNIST training and testing takes less than 15 minutes with an acceptable recognition rate. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Deep Learning / Convolutional Neural Network / CUDA / GPGPU |
Paper # | NC2014-3,IBISML2014-3 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 2014/6/18(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Neurocomputing (NC) |
---|---|
Language | ENG |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Implementation of Deep Neural Network on Image Classification |
Sub Title (in English) | |
Keyword(1) | Deep Learning |
Keyword(2) | Convolutional Neural Network |
Keyword(3) | CUDA |
Keyword(4) | GPGPU |
1st Author's Name | Rui ZHONG |
1st Author's Affiliation | The Graduate School of Library, Information and Media Studies, University of Tsukuba() |
2nd Author's Name | Taro TEZUKA |
2nd Author's Affiliation | The Graduate School of Library, Information and Media Studies, University of Tsukuba |
Date | 2014-06-25 |
Paper # | NC2014-3,IBISML2014-3 |
Volume (vol) | vol.114 |
Number (no) | 104 |
Page | pp.pp.- |
#Pages | 7 |
Date of Issue |