search

Jia1018 / ResGEM

Mesh Denoising with Graph Convolutional Networks
9 stars 0 forks source link

How to preprosess noisy mesh and generate denoisy mesh #3

Open zhuzhiwei99 opened 1 month ago

zhuzhiwei99 commented 1 month ago

How to preprosess the noisy mesh to get the input datas image

I run the cmd python infer.py --testset ./data/test --data_type Synthetic --graph_type DENSE_FACE --resume ./ckpt/Synthetic_nn.pth --save_dir ./data/results And I get the following result. What's the meaning of the result? Are they denoisy normals? How can I use these results to get denoisy mesh. image

Hope author can provide necessary codes to get full results. thanks!

Jia1018 commented 1 month ago

Yes, the results are the denoised normals of these models. The data preprocessing process is implemented in C++ code using the easy3D toolkit. The compilation and execution of the code are relatively complex, so the code is not provided. If you need the complete denoising results on the public dataset, you can download them directly from the following link: https://drive.google.com/drive/folders/1cV3offz01vJhnpo8ECGSK-OYQKzT3OXJ?usp=drive_link. I hope this information proves helpful in your research, and please feel free to reach out if you have any further questions or require additional assistance.