Skip to content

Latest commit

 

History

History
23 lines (20 loc) · 956 Bytes

README.md

File metadata and controls

23 lines (20 loc) · 956 Bytes

Pytorch Implementation of the Attacks on PointNet++

Semantic Segmentation

Data Preparation

Download 3D indoor parsing dataset (S3DIS) here and save in data/Stanford3dDataset_v1.2_Aligned_Version/.

cd data_utils
python collect_indoor3d_data.py

Processed data will save in data/stanford_indoor3d/.

Run

python NB_nontarget_test_semseg.py --data_path <data_path>

Great thanks for the PointNet++. The attack code is implemented based on the adversarial-attacks-pytorch.

Reference:

halimacc/pointnet3
fxia22/pointnet.pytorch
charlesq34/PointNet
charlesq34/PointNet++