3D-MPA: Multi Proposal Aggregation for 3D Semantic Instance Segmentation

IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020.

Introduction

This repository is code release for our CVPR 2020 paper (ArXiv Paper)

While we prepare the code release, have a look at the interactive demo on our project page and the teaser video:

• Project Page
• YouTube Video

Predictions on ScanNet

We provide the predictions of 3D-MPA over all three ScanNet splits: 3D-MPA ScanNet Predictions
The format of the predictions is the same as used for evaluation on the ScanNet online benchmark, as described here.

Citation

If you find our work useful in your research, please consider citing:

@inproceedings{Engelmann20CVPR,
  title = {{3D-MPA: Multi Proposal Aggregation for 3D Semantic Instance Segmentation}},
  author = {Engelmann, Francis and Bokeloh, Martin and Fathi, Alireza and Leibe, Bastian and Nie{\ss}ner, Matthias},
  booktitle = {{IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}},
  year = {2020}
}