Name
Abstract | ||
|---|---|---|
We propose building-scale virtual reality (VR), a real-world extension method different from augmented reality (AR), that uses automatically generated indoor 3D point cloud maps. To make the entire indoor area a pose tracking area, we attach an RGB-D camera to a VR headset, collect data with it, and use deep learning to build a model learned from the data. This method is more accurate than the conventional RGB-D SLAM method. Furthermore, to modify the VR space according to the purpose, segmentation and replacement of the 3D point cloud are performed. This is hard to do in AR, but it is essential technology for VR to be used in actual real-world work. We describe a disaster simulation including virtual evacuation drills and virtual work environments as application examples. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1109/MIPR.2019.00044 | 2019 IEEE Conference on Multimedia Information Processing and Retrieval (MIPR) |
Keywords | Field | DocType |
Building-Scale Virtual Reality, Augmented Reality, 3D Scene Reconstruction, 3D Pose Estimation, 3D Scene Modification, Disaster Simulation | Headset,Computer vision,Virtual reality,Extension method,Computer science,3D pose estimation,Augmented reality,Artificial intelligence,Deep learning,Virtual work,Point cloud | Conference |
ISBN | Citations | PageRank |
978-1-7281-1198-8 | 1 | 0.37 |
References | Authors | |
0 | 4 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Katashi Nagao | 1 | 559 | 196.30 |
| Menglong Yang | 2 | 109 | 10.49 |
| Xu Cao | 3 | 5 | 8.07 |
| Yusuke Miyakawa | 4 | 2 | 0.74 |