| Presentation | 2017-02-20 A parallel computation method of super resolution using convolutional neural networks Yusuke Sugawara, Sayaka Shiota, Hitoshi Kiya, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | An acceleration method for interpolation-based super resolutionmethods using convolutional neural networks, represented by SRCNN andVDSR is proposed in this article. In the proposed method, estimatedpixels are classified into some types and then high resolution imagesare directly generated from low resolution ones by using optimizedconvolutional neural networks for each type. It allows us to adaptsmaller filter sizes to convolutional neural networks thanconventional ones, so that the computational complexity can bereduced. A number of experiments are carried out to demonstrate thatthe proposed method can be applied to various interpolation-basedsuper resolution methods using convolutional neural networks. Theeffectiveness of the proposed method for multiple upscale factors isalso confirmed. In addition, the proposed method outperformsconventional ones in terms of the quality of estimated images under some conditions. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Super Resolution / Convolutional Neural Network / Parallel Computation |
| Paper # | ITS2016-44,IE2016-102 |
| Date of Issue | 2017-02-13 (ITS, IE) |
| Conference Information | |
| Committee | IE / ITS / ITE-AIT / ITE-HI / ITE-ME / ITE-MMS / ITE-CE |
|---|---|
| Conference Date | 2017/2/20(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Hokkaido Univ. |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Image Processing, etc. |
| Chair | Seishi Takamura(NTT) / Tomotaka Nagaosa(Kanto Gakuin Univ.) / Tokiichiro Takahashi(TDU) / Masayuki Sato(Univ. of Kitakyushu) / Miki Haseyama(Hokkaido Univ.) / Eiichi Miyashita(NHK) / Koji Minami(Mitsubishi Electric Corp.) |
| Vice Chair | Takayuki Hamamoto(Tokyo Univ. of Science) / Atsuro Ichigaya(NHK) / Masahiro Fujii(Utsunomiya Univ.) / Tomotaka Wada(Kansai Univ.) / / / Norio Tagawa(Tokyo Metropolitan Univ.) |
| Secretary | Takayuki Hamamoto(NTT) / Atsuro Ichigaya(Chiba Inst. of Tech.) / Masahiro Fujii(Meiji Univ.) / Tomotaka Wada(AIST) / (Tokyo Polytechnic Univ.) / (NICT) / Norio Tagawa(NTT) / (NHK) / (NHK) |
| Assistant | Kei Kawamura(KDDI R&D Labs.) / Keita Takahashi(Nagoya Univ.) / Tetsuya Manabe(Saitama Univ.) / Yanlei Gu(Univ. of Tokyo) / Koichiro Hashiura(Akita Pref. Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Image Engineering / Technical Committee on Intelligent Transport Systems Technology / Technical Group on Artistic Image Technology / Technical Group on Human Inormation / Technical Group on Media Engineering / Technical Group on Multi-media Storage / Technical Group on Consumer Electronics |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | A parallel computation method of super resolution using convolutional neural networks |
| Sub Title (in English) | |
| Keyword(1) | Super Resolution |
| Keyword(2) | Convolutional Neural Network |
| Keyword(3) | Parallel Computation |
| 1st Author's Name | Yusuke Sugawara |
| 1st Author's Affiliation | Tokyo Metro. Univ.(Tokyo Metro. Univ.) |
| 2nd Author's Name | Sayaka Shiota |
| 2nd Author's Affiliation | Tokyo Metro. Univ.(Tokyo Metro. Univ.) |
| 3rd Author's Name | Hitoshi Kiya |
| 3rd Author's Affiliation | Tokyo Metro. Univ.(Tokyo Metro. Univ.) |
| Date | 2017-02-20 |
| Paper # | ITS2016-44,IE2016-102 |
| Volume (vol) | vol.116 |
| Number (no) | ITS-463,IE-464 |
| Page | pp.pp.13-18(ITS), pp.13-18(IE), |
| #Pages | 6 |
| Date of Issue | 2017-02-13 (ITS, IE) |