BUFF dataset evaluation

[MoyGcc]

Hi,

thanks for your amazing work. Could you please share the code snippet for the rendering and evaluation of the BUFF dataset? Many thanks in advance!

[CrisHY1995]

The rendering code of BUFF Dataset has been supplied in this Repository(https://github.com/CrisHY1995/StereoPIFu_Code/blob/main/GenEvalData.py#L105). Please refer to the introduction of project page(https://github.com/CrisHY1995/StereoPIFu_Code) and run the command ``` python GenEvalData.py --color_mesh_path="TempData/SampleData/normalized_mesh_0089.off" --save_dir="./TempData/ColorMesh" --save_postfix="color" ``` Note: the mesh used for rendering needs to be scaled and let it located in a specific bounding box. As to evaluation, because the related code is scattered, it might not be easy for you to use. Instead, we summarize the main steps as follows: 1. Register predicted mesh with GT mesh using this method (https://github.com/yaoyx689/Fast-Robust-ICP ) 2. Compute the Predict-to-GT/GT-to-Predict/Chamfer distance using pymesh library (https://pymesh.readthedocs.io/en/latest/api_geometry_processing.html), the kernel function is “pymesh.distance_to_mesh(mesh, pts, engine='geogram')” Best Wishes. Yang. title: [CrisHY1995/StereoPIFu_Code] BUFF dataset evaluation (#1) Hi, thanks for your amazing work. Could you please share the code snippet for the rendering and evaluation of the BUFF dataset? Many thanks in advance! ― You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#1>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ALYN3WKHHKJ6YLD36VTUSMTTLVHTVANCNFSM437M5B6Q>.

[CrisHY1995]

PS:

1. Before registering, you need to delete free parts of the BUFF data except the main body.
2. Calculate the registration scale based on the y-axis range of the prediction result and the y-axis range of GT.

[MoyGcc]

PS:

1. Before registering, you need to delete free parts of the BUFF data except the main body.
2. Calculate the registration scale based on the y-axis range of the prediction result and the y-axis range of GT.

Many thanks!