| Presentation | 2019-03-05 Adaptive Data Compression Using Deep Learning Executable on Edge Device Masatoshi Sekine, Satoshi Ikada, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Efficient collection of sensor data is important for sensing the real world and utilizing it in various applications. Generally, a large amount of sensor data is required to analyze long-term and wide-range data with high precision in applications. However, at the edge device, waste of radio resources and power due to an increase in the amount of data transmission may occur. Therefore, it is necessary to reduce the amount of data without degrading the information amount of the original data as much as possible. In this paper, we propose a method to set the data transmission efficiency and optimize it to estimate the compression ratio of the compressed sensing. It is executable at the edge device dynamically and directly, in order to improve data transmission efficiency. Generally, optimization control generally requires a large amount of calculation amount, but by using the learning model of supervised learning or reinforcement learning in deep learning learned in advance in the proposed method, the search for optimal compression ratio can be performed with less processing load. As a result of performance evaluation, using the data generated by an acceleration sensor placed on the bridge, the proposed method can reduce the computational load for estimating the optimum compression ratio, while reducing both compression ratio and reconstruction error. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | sensor networks / compressed sensing / edge device / deep learning / supervised learning / reinforcement learning |
| Paper # | NS2018-293 |
| Date of Issue | 2019-02-25 (NS) |
| Conference Information | |
| Committee | IN / NS |
|---|---|
| Conference Date | 2019/3/4(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Okinawa Convention Center |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | General |
| Chair | Takuji Kishida(NTT-AT) / Yoshikatsu Okazaki(NTT) |
| Vice Chair | Kenji Ishida(Hiroshima City Univ.) / Akihiro Nakao(Univ. of Tokyo) |
| Secretary | Kenji Ishida(KDDI Research) / Akihiro Nakao(KDDI Research) |
| Assistant | / Kenichi Kashibuchi(NTT) |
| Paper Information | |
| Registration To | Technical Committee on Information Networks / Technical Committee on Network Systems |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Adaptive Data Compression Using Deep Learning Executable on Edge Device |
| Sub Title (in English) | |
| Keyword(1) | sensor networks |
| Keyword(2) | compressed sensing |
| Keyword(3) | edge device |
| Keyword(4) | deep learning |
| Keyword(5) | supervised learning |
| Keyword(6) | reinforcement learning |
| 1st Author's Name | Masatoshi Sekine |
| 1st Author's Affiliation | Oki Electric Industry Co., Ltd.(OKI) |
| 2nd Author's Name | Satoshi Ikada |
| 2nd Author's Affiliation | Oki Electric Industry Co., Ltd.(OKI) |
| Date | 2019-03-05 |
| Paper # | NS2018-293 |
| Volume (vol) | vol.118 |
| Number (no) | NS-465 |
| Page | pp.pp.575-580(NS), |
| #Pages | 6 |
| Date of Issue | 2019-02-25 (NS) |