Presentation | 2017-03-21 Three-dimensional shape retrieval based in voxel representation using LSTM and CNN Ryo Miyagi, Masaki Aono, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | In recent years, with the spread of 3D scanners, VR headsets, etc., there is an increasing demand for recognition and retrieval for 3D contents. Furthermore, by applying deep learning to 3-dimensional objects, research on shape classification, shape search and the like are actively conducted. In this paper, we represent our model with binary voxels and propose a new method for searching similar 3D shape models using deep learning. Specifically, we employ a Convolutional Neural Network (CNN) to represent a 2D slice extracting from 3D binary voxels, and a Long Short-Term Memory (LSTM) to represent 2D slices as time-series connected features showing a given 3D shape. As a result of the experiment, we could improve the performance of our method against a baseline method which is based on 3DCNN. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Deep Learning / 3D / LSTM / CNN / Shape Retrieval / Binary Voxel |
Paper # | BioX2016-67,PRMU2016-230 |
Date of Issue | 2017-03-13 (BioX, PRMU) |
Conference Information | |
Committee | PRMU / BioX |
---|---|
Conference Date | 2017/3/20(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | Eisaku Maeda(NTT) / Masakatsu Nishigaki(Shizuoka Univ.) |
Vice Chair | Seiichi Uchida(Kyushu Univ.) / Hironobu Fujiyoshi(Chubu Univ.) / Akira Otsuka(AIST) / Hiroshi Takano(Toyama Pref. Univ.) |
Secretary | Seiichi Uchida(Kyoto Univ.) / Hironobu Fujiyoshi(NTT) / Akira Otsuka(NEC) / Hiroshi Takano(AIST) |
Assistant | Masaki Oonishi(AIST) / Takuya Funatomi(NAIST) / Masatsugu Ichino(Univ. of Electro-Comm.) / Naoyuki Takada(Secom) / Takahiro Aoki(Fujitsu Labs.) |
Paper Information | |
Registration To | Technical Committee on Pattern Recognition and Media Understanding / Technical Committee on Biometrics |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Three-dimensional shape retrieval based in voxel representation using LSTM and CNN |
Sub Title (in English) | |
Keyword(1) | Deep Learning |
Keyword(2) | 3D |
Keyword(3) | LSTM |
Keyword(4) | CNN |
Keyword(5) | Shape Retrieval |
Keyword(6) | Binary Voxel |
1st Author's Name | Ryo Miyagi |
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 | 2017-03-21 |
Paper # | BioX2016-67,PRMU2016-230 |
Volume (vol) | vol.116 |
Number (no) | BioX-527,PRMU-528 |
Page | pp.pp.203-208(BioX), pp.203-208(PRMU), |
#Pages | 6 |
Date of Issue | 2017-03-13 (BioX, PRMU) |