Zhu Yuefeng

I followed the COLMAP tutorial to produce a custom dataset. It has same structure as in 'Quickstart' chapter of tutorial. However, your preprocess shell script cannot work well with that,...

update: when trying manually make custom trainjng data, find bad support in hard-code style When work manually, step 1 and 2 go well. In step 3 generate cache, the 'data_generation.sh'...

update: worked out. I changed downsample in config yaml, and paths became matching well.

Hi, I was on holiday this weekend~~ Now I am back to code. In all previous experiments I used default 'tet_res'=90. By changing it to 100 I see little improvement...

Update: I made a super-high resolution *.npz file, but GPU memory is not large enough to support. Details: I installed quartet, and then followed your instructions in `data/generate_tets.py` to produce...

Another question: why we apply quartet here? I read about quartet and it claims to be a method for approximating some given mesh with a uniform mesh. According to example...

An experiment with --tet_res=64 also cost 11G memory at peak. I am confused about memory and resolution.

> @zhu-yuefeng Hello, may I ask you the way you solved the hard-code problem? I am truly sorry that I can hardly help. After failing to get satisfying result on...

Same case. Details: - Env: Windows10 + python 3.10 + Cuda 12.0 (Driver 528.24) + cudNN 8.8.1 - Action: (PowerScript) .\install.bat - Output: - ... - Checking cuDNN... cuDNN is...