about pre-trained model
Closed this issue · 1 comments
Hi, i'm studying about 3D vision via point cloud.
in point clean net, i want to run your pre trained model but there are two files each models "~model.pth" and "~params.pth".
i tried to run for my point data sets (69444, 6)
model_outlier = torch.load("/content/PointCleanNetOutliers_model.pth")
model_outlier.eval()
model_denoise = torch.load("/content/Denoising/PointCleanNet_model.pth")
model_denoise.eval()
model_outlier = torch.load("/content/PointCleanNetOutliers_params.pth")
model_outlier.eval()
model_denoise = torch.load("/content/Denoising/PointCleanNet_params.pth")
model_denoise.eval()
model_outlier = torch.load("/content/PointCleanNetOutliers_model.pth")
model_outlier.load_state_dict(torch.load("/content/PointCleanNetOutliers_params.pth"))
model_outlier.eval()
model_denoise = torch.load("/content/Denoising/PointCleanNet_model.pth")
model_denoise.load_state_dict(torch.load("/content/PointCleanNet_params.pth"))
model_denoise.eval()
Nothing has been done...
can you tell me how?
i'm a beginner 🥲...
Hi @eunseoday
Have you tried running the following in a terminal (this code is also in the read me)?
For outlier removal:
cd outliers_removal
mkdir results
python eval_pcpnet.py
For denoising:
cd noise_removal
mkdir results
./run.sh
Normally these run the pretrained model. If you want to run them separately and this works, you could then copy code from the eval_pcpnet.py files?