3D object completion via class-conditional generative adversarial network

Yu Chieh Chen, Daniel Stanley Tan, Wen-Huang Cheng, Kai Lung Hua*

*Corresponding author for this work

研究成果: Conference contribution同行評審

6 引文 斯高帕斯(Scopus)

摘要

Many robotic tasks require accurate shape models in order to properly grasp or interact with objects. However, it is often the case that sensors produce incomplete 3D models due to several factors such as occlusion or sensor noise. To address this problem, we propose a semi-supervised method that can recover the complete the shape of a broken or incomplete 3D object model. We formulated a hybrid of 3D variational autoencoder (VAE) and generative adversarial network (GAN) to recover the complete voxelized 3D object. Furthermore, we incorporated a separate classifier in the GAN framework, making it a three player game instead of two which helps stabilize the training of the GAN as well as guides the shape completion process to follow the object class labels. Our experiments show that our model produces 3D object reconstructions with high-similarity to the ground truth and outperforms several baselines in both quantitative and qualitative evaluations.

原文English
主出版物標題MultiMedia Modeling - 25th International Conference, MMM 2019, Proceedings
編輯Benoit Huet, Ioannis Kompatsiaris, Stefanos Vrochidis, Vasileios Mezaris, Wen-Huang Cheng, Cathal Gurrin
發行者Springer Verlag
頁面54-66
頁數13
ISBN(列印)9783030057152
DOIs
出版狀態Published - 1 一月 2019
事件25th International Conference on MultiMedia Modeling, MMM 2019 - Thessaloniki, Greece
持續時間: 8 一月 201911 一月 2019

出版系列

名字Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
11296 LNCS
ISSN(列印)0302-9743
ISSN(電子)1611-3349

Conference

Conference25th International Conference on MultiMedia Modeling, MMM 2019
國家Greece
城市Thessaloniki
期間8/01/1911/01/19

指紋 深入研究「3D object completion via class-conditional generative adversarial network」主題。共同形成了獨特的指紋。

引用此