Danpeng Chen

Thank you for your attention to our work. The output includes a debug folder. Could you please provide the results from the last iterations to help us pinpoint the issue?

可视化看着还正常。您有按照我们的脚步方法预处理tnt数据吗？可以将gt和生成的mesh一起可视化一下，确认坐标系是不是对齐？

Hello, are there any error messages? During training, you can remove the --quiet parameter to allow error information to be printed.

Hello, we use the Open3D tool to display the mesh and normals. After opening the mesh in Open3D, press the shortcut ctrl+9 to toggle the normal map, and press 9...

Hello, the result is quite poor. Are your data preprocessing and training parameters consistent with the README?

Extracting 1/5 of the images results in overly sparse data, making high-quality reconstruction impossible. You can refer to the script scripts/run_tnt.py, and add the argument ‘--data_device cpu’ to cache the...

Hi, could you show your folder structure? I'm guessing you need to execute the following commands: python scripts/preprocess/convert.py --data_path /opt/data/private/chy/data

You can refer to the **Custom Dataset** data processing instructions in the README. For data preprocessing, make sure the data paths are correct. When executing the command, you only need...

TSDF may cause color distortion due to resolution and color averaging in different directions. To achieve better textures, it is recommended to perform texture optimization when a mesh is available,...

Hello, the plane_depth is derived from rendered_normal and rendered_distance. The formula in the code is theoretically consistent with the one in the paper. For simplicity of implementation, we directly use...