Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

NPM3D dataset #40

Closed
xbb2017 opened this issue Nov 11, 2019 · 2 comments
Closed

NPM3D dataset #40

xbb2017 opened this issue Nov 11, 2019 · 2 comments

Comments

@xbb2017
Copy link

xbb2017 commented Nov 11, 2019

Hi Hugues, did you use reflectance feature during training in NPM3D dataset? Thank you.

@HuguesTHOMAS
Copy link
Owner

HuguesTHOMAS commented Nov 11, 2019

Hi @xbb2017,

No I did not. If you want to use it, you might need to correct the values beforehand. The raw reflectance need to be calibrated to actually mean something.

Best,
Hugues

@xbb2017
Copy link
Author

xbb2017 commented Nov 13, 2019

Hi @xbb2017,

No I did not. If you want to use it, you might need to correct the values beforehand. The raw reflectance need to be calibrated to actually mean something.

Best,
Hugues

Thank you so much Hugues. I tried to train NPM3D dataset as you suggested (#15 (comment)) under CUDA9.2, tensorflow 1.12.3 and GeForce GTX 1080 Ti. The output loss still became NaN after several epochs. It seems not the internal bug of CUDA10. I found this bug happened randomly, sometimes happened even after 500 epochs.... Another weird thing is that this bug never happen when I training Semantic3D dataset.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants