三维重建-基于3D扫描的大规模场景补全+语义分割-优质项目实战.zip

# ScanComplete ScanComplete is a data-driven approach which takes an incomplete 3D scan of a scene as input and predicts a complete 3D model, along with per-voxel semantic labels. [<img src="images/teaser_mesh.jpg">] ## Code ### Installation: Training is implemented with [TensorFlow](https://www.tensorflow.org/). This code is tested under TF1.3 and Python 2.7 on Ubuntu 16.04. ### Training: * See `run_train.sh` for calling the training (will need to provide a path to the train data). * Trained models: [models.zip](http://kaldir.vc.in.tum.de/adai/ScanComplete/models.zip) ### Testing: * See `run_complete_scans_hierarchical.sh` for testing on partial scans (needs paths to test data and model).