| Presentation | 2017-10-12 Online Human Action Detection using Deep Spatio-temporal Transformation Yukihide Takagaki, Masaki Aono, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In this research, we describe online human action detection using Convolutional Neural Network which inputs skeleton data of multiple frames. We propose a network that transforms skeleton data into a certain image by the Fully connected layer and uses the image as input to the 3D convolution layer. Furthermore, we output the probability of the action class of the frame, and apply an average filter multiple times to the probability among frames for improvement of performance. In the experiment, we compared online human action detection accuracy between conventional method and proposed method using Online Action Detection Dataset consisting of skeleton data, image data, and depth data of human action. As a result, we could improve the performance of our method against conventional methods. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Skeleton / Online Action Detection / Transform / CNN / Temporal Localization |
| Paper # | PRMU2017-68 |
| Date of Issue | 2017-10-05 (PRMU) |
| Conference Information | |
| Committee | PRMU |
|---|---|
| Conference Date | 2017/10/12(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Shinichi Sato(NII) |
| Vice Chair | Hironobu Fujiyoshi(Chubu Univ.) / Yoshihisa Ijiri(Omron) |
| Secretary | Hironobu Fujiyoshi(AIST) / Yoshihisa Ijiri(NAIST) |
| Assistant | Masato Ishii(NEC) / Yusuke Sugano(Osaka Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Pattern Recognition and Media Understanding |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Online Human Action Detection using Deep Spatio-temporal Transformation |
| Sub Title (in English) | |
| Keyword(1) | Skeleton |
| Keyword(2) | Online Action Detection |
| Keyword(3) | Transform |
| Keyword(4) | CNN |
| Keyword(5) | Temporal Localization |
| 1st Author's Name | Yukihide Takagaki |
| 1st Author's Affiliation | Toyohashi University of Technology(TUT) |
| 2nd Author's Name | Masaki Aono |
| 2nd Author's Affiliation | Toyohashi University of Technology(TUT) |
| Date | 2017-10-12 |
| Paper # | PRMU2017-68 |
| Volume (vol) | vol.117 |
| Number (no) | PRMU-238 |
| Page | pp.pp.31-35(PRMU), |
| #Pages | 5 |
| Date of Issue | 2017-10-05 (PRMU) |