Presentation | 2018-12-14 Training of Traffic Sign Detector and Classifier Using Synthetic Road Scenes Akira Sekizawa, Katsuto Nakajima, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | This paper proposes a method of providing an end-to-end object recognition system based on deep learning that uses synthetically generated road scenes as recognition training images in place of actual images of road signs. Conventional training image generation methods often generate only small images that include just the traffic sign part and are not capable of training end-to-end object recognition systems using entire scenes as the training images. This paper proposes a method of synthetically generating road scenes as end-to-end object recognition system training data; the system shows that generating scenes considering contextual information around traffic signs effectively improves precision. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Traffic Sign Recognition / Object Detection / Synthetic Data / Data Augmentation |
Paper # | PRMU2018-89 |
Date of Issue | 2018-12-06 (PRMU) |
Conference Information | |
Committee | PRMU |
---|---|
Conference Date | 2018/12/13(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 | Yoshihisa Ijiri(Omron) / Toru Tamaki(Hiroshima Univ.) |
Secretary | Yoshihisa Ijiri(NEC) / Toru Tamaki(Osaka Univ.) |
Assistant | Go Irie(NTT) / Yoshitaka Ushiku(OSX) |
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) | Training of Traffic Sign Detector and Classifier Using Synthetic Road Scenes |
Sub Title (in English) | |
Keyword(1) | Traffic Sign Recognition |
Keyword(2) | Object Detection |
Keyword(3) | Synthetic Data |
Keyword(4) | Data Augmentation |
1st Author's Name | Akira Sekizawa |
1st Author's Affiliation | Tokyo Denki University(TDU) |
2nd Author's Name | Katsuto Nakajima |
2nd Author's Affiliation | Tokyo Denki University(TDU) |
Date | 2018-12-14 |
Paper # | PRMU2018-89 |
Volume (vol) | vol.118 |
Number (no) | PRMU-362 |
Page | pp.pp.73-78(PRMU), |
#Pages | 6 |
Date of Issue | 2018-12-06 (PRMU) |