4DenoiseNet: Adverse Weather Denoising from Adjacent Point Clouds, Publication

Citation:

@ARTICLE{9976208, author={Seppanen, Alvari and Ojala, Risto and Tammi, Kari}, journal={IEEE Robotics and Automation Letters}, title={4DenoiseNet: Adverse Weather Denoising from Adjacent Point Clouds}, year={2022}, volume={}, number={}, pages={1-8}, doi={10.1109/LRA.2022.3227863}}

SnowyKITTI-dataset:

Download

Train:

cd networks
./train.sh -d root/snowyKITTI/dataset/ -a fourdenoisenet.yml -l /your/log/folder/ -c 0

Infer (pretrained model -m root/logs/2023-1-17-08:49/):

cd networks/train/tasks/semantic
python3 infer.py -d root/toy_snowyKITTI/dataset/ -m root/logs/2023-1-17-08:49/ -l /your/predictions/folder/ -s test
(-s = split)

Evaluate:

cd networks/train/tasks/semantic
python3 snow_evaluate_iou.py -d root/toy_snowyKITTI/dataset/ -dc root/networks/train/tasks/semantic/config/labels/snowy-kitti.yaml -p /your/predictions/folder/ -s test
(-s = split)

Visualize:

cd utils
python3 snow_visualize.py -d root/toy_snowyKITTI/dataset/ -c root/networks/train/tasks/semantic/config/labels/snowy-kitti.yaml -p /your/predictions/folder/ -s 22
(-s = sequence)

Thanks to SalsaNext for providing some of the code!