Dataset Open Access
3DFacilities
Czerniawski,Thomas;
Leite, Fernanda
Please visit the 3DFacilities website, thomasczerniawski.com/3dfacilities
3DFacilities is an annotated RGB-D data-set of building facilities. It currently contains over 25,000 individual frames and 110 scene reconstructions.
The data-set was created for training machine learning algorithms to perform semantic segmentation. Classes included in the data-set are: furniture, door, wall, floor, window, ceiling, column, beam, stairs, railing, light fixture, elevator, plumbing, duct, diffuser, sprinkler, cable tray, and conduit
Files
(3.8 GB)
| Name | Size | |
|---|---|---|
|
3DFacilities_Complete.zip
md5:e103c55ef7b301739bad8b364a5f0724 |
3.8 GB | Download |
|
ReadMe.txt
md5:7c2e78b91b904c680592292718e04859 |
2.5 kB | Download |
850
229
views
downloads
| All versions | This version | |
|---|---|---|
| Views | 850 | 392 |
| Downloads | 229 | 159 |
| Data volume | 408.4 GB | 337.0 GB |
| Unique views | 767 | 357 |
| Unique downloads | 169 | 118 |
Indexed in
- Publication date:
- June 18, 2018
- DOI:
-
- Keyword(s):
- Computer Vision Deep Learning Image Classification Semantic Segmentation Building Systems Building Information Modeling Building Information Modelling Architecture Facility Management Images RGB-D Structure IO Range Camera Range Images Depth Images Depth Camera 3D Reconstruction 3D Imaging 3D Images Machine Learning
- Grants:
-
National Science Foundation:
- Collaborative Research: Living Building Information Model (BIM): A Layered Approach for Automatic and Continuous Built Environment Model Update (1562438)
- License (for files):
- Creative Commons Attribution Non Commercial 4.0 International