Skip to content

Latest commit

 

History

History
38 lines (26 loc) · 1010 Bytes

preprocess_nyu.md

File metadata and controls

38 lines (26 loc) · 1010 Bytes
Raw

Please get the nyuv2 images from here

Download the proposals and object data for nyu

tar -xf nyuv2.tar.gz nyuv2
cd nyuv2

Download the splits. Please make sure the splits are in the nyuv2 dir

tar -xf nyu_splits.tar.gz splits

Move NYU images to nyuv2 dir.

Preprocess the NYU data. Create the mat files using the script

python voxelize_objects.py

Please consider citing if you use this for your reserach

@inproceedings{gupta2014learning,
  title={Learning rich features from RGB-D images for object detection and segmentation},
  author={Gupta, Saurabh and Girshick, Ross and Arbel{\'a}ez, Pablo and Malik, Jitendra},
  booktitle={European Conference on Computer Vision},
  pages={345--360},
  year={2014},
  organization={Springer}
}