Semantic Segmentation of point clouds using range images.

Developed by Andres Milioto, Jens Behley, Ignacio Vizzo, and Cyrill Stachniss

Examples of segmentation results from SemanticKITTI dataset:

This code provides code to train and deploy Semantic Segmentation of LiDAR scans, using range images as intermediate representation. The training pipeline can be found in /train. We will open-source the deployment pipeline soon.

• squeezeseg
• squeezeseg + crf
• squeezesegV2
• squeezesegV2 + crf
• darknet21
• darknet53
• darknet53-1024
• darknet53-512

To enable kNN post-processing, just change the boolean value to True in the arch_cfg.yaml file parameter, inside the model directory.

These are the predictions for the train, validation, and test sets. The performance can be evaluated for the training and validation set, but for test set evaluation a submission to the benchmark needs to be made (labels are not public).

No post-processing:

• squeezeseg
• squeezeseg + crf
• squeezesegV2
• squeezesegV2 + crf
• darknet21
• darknet53
• darknet53-1024
• darknet53-512

With k-NN processing:

• squeezeseg
• squeezesegV2
• darknet53
• darknet21
• darknet53-1024
• darknet53-512

Copyright 2019, Andres Milioto, Jens Behley, Cyrill Stachniss. University of Bonn.

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

The pretrained models with a specific dataset maintain the copyright of such dataset.

If you use our framework, model, or predictions for any academic work, please cite the original paper, and the dataset.

@inproceedings{milioto2019iros,
  author    = {A. Milioto and I. Vizzo and J. Behley and C. Stachniss},
  title     = {{RangeNet++: Fast and Accurate LiDAR Semantic Segmentation}},
  booktitle = {IEEE/RSJ Intl.~Conf.~on Intelligent Robots and Systems (IROS)},
  year      = 2019,
  codeurl   = {https://github.com/PRBonn/lidar-bonnetal},
  videourl  = {https://youtu.be/wuokg7MFZyU},
}

@inproceedings{behley2019iccv,
  author    = {J. Behley and M. Garbade and A. Milioto and J. Quenzel and S. Behnke and C. Stachniss and J. Gall},
  title     = {{SemanticKITTI: A Dataset for Semantic Scene Understanding of LiDAR Sequences}},
  booktitle = {Proc. of the IEEE/CVF International Conf.~on Computer Vision (ICCV)},
  year      = {2019}
}