search

qq456cvb / Point-Transformers

Point Transformers
MIT License
609 stars 102 forks source link

About part segmentation performance #18

Open RESET0 opened 2 years ago

RESET0 commented 2 years ago

I have trained the part segmentation part following your train_partseg.py. And the performance is not well.Is there anything I need to pay attention to ? Best accuracy is: 0.64983, Best class avg mIOU is: 0.36181, Best inctance avg mIOU is: 0.50138

zhangdongwei1998 commented 2 years ago

Hello, I got this error ‘RuntimeError: CUDA error: CUBLAS_STATUS_NOT_INITIALIZED when calling cublasCreate(handle)’ when I was doing the segmentation task, why is this?thanks

zhangdongwei1998 commented 2 years ago

So I can't run this segmentation code correctly.Do you know why?thank you.

qq456cvb commented 2 years ago

@RESET0 Do you experience a large gap between training/testing, and is the training accuracy also very low?

Lizhinwafu commented 2 years ago

I have trained the part segmentation part following your train_partseg.py. And the performance is not well.Is there anything I need to pay attention to ? Best accuracy is: 0.64983, Best class avg mIOU is: 0.36181, Best inctance avg mIOU is: 0.50138

Can you provide the test file? I want to visualize the generated prediction file.