| Presentation | 2022-01-27 Proposal and evaluation of 3D-point object estimation method based on probability space representation Hiroaki Sato, Shin'ichi Arakawa, Masayuki Murata, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | New network services are expected to emerge using real spatial information in remote areas. For the advancement of services, it is important to understand real space from real spatial information, and for this purpose, object identification technology using machine learning is being introduced. In this paper, we have tried to represent real space as a field of probabilistic superposition of objects, instead of directly identifying real space information using machine learning techniques. We obtain neighboring information based on the positional relationships of objects in real space from a dataset used by the machine learning. Then, we propose a method of estimating objects by adding empirical knowledge information of real space to the object identification results. Our results show that predictions are changed by adapted empirical knowledge information and the estimation accuracy is improved when the confidence by the machine learning is not high. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | point cloud / random field representation / object detection / deep learning / semantic segmentation / Bayesian estimation |
| Paper # | CQ2021-83 |
| Date of Issue | 2022-01-20 (CQ) |
| Conference Information | |
| Committee | CQ |
|---|---|
| Conference Date | 2022/1/27(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Kanazawa（Ishikawa Pref.） |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | AR/VR, Broadcasting Service, Video/Voice Services Quality, High Realistic, User Behavior/Psychology, User Experience, Media Quality, Network Quality and QoS Control, Networks and Communications at Disaster, User Behavior, Machine Learning, Video Communication, etc. |
| Chair | Jun Okamoto(NTT) |
| Vice Chair | Takefumi Hiraguri(Nippon Inst. of Tech.) / Gou Hasegawa(Tohoku Univ.) |
| Secretary | Takefumi Hiraguri(NTT) / Gou Hasegawa(Ritsumeikan Univ.) |
| Assistant | Yoshiaki Nishikawa(NEC) / Ryoichi Kataoka(KDDI Research) / Kimiko Kawashima(NTT) |
| Paper Information | |
| Registration To | Technical Committee on Communication Quality |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Proposal and evaluation of 3D-point object estimation method based on probability space representation |
| Sub Title (in English) | |
| Keyword(1) | point cloud |
| Keyword(2) | random field representation |
| Keyword(3) | object detection |
| Keyword(4) | deep learning |
| Keyword(5) | semantic segmentation |
| Keyword(6) | Bayesian estimation |
| 1st Author's Name | Hiroaki Sato |
| 1st Author's Affiliation | Osaka University(Osaka Univ.) |
| 2nd Author's Name | Shin'ichi Arakawa |
| 2nd Author's Affiliation | Osaka University(Osaka Univ.) |
| 3rd Author's Name | Masayuki Murata |
| 3rd Author's Affiliation | Osaka University(Osaka Univ.) |
| Date | 2022-01-27 |
| Paper # | CQ2021-83 |
| Volume (vol) | vol.121 |
| Number (no) | CQ-357 |
| Page | pp.pp.39-44(CQ), |
| #Pages | 6 |
| Date of Issue | 2022-01-20 (CQ) |