Ray distribution loss (sigma_loss) is not used?

[niujinshuchong]

Hi, thanks for sharing the great work.

I run with the default config file and it works well. However, it use the MSE loss for depth supervision and the ray distribution loss is not used by default. I try to enable it by setting sigma_loss = True in the config file but it raise an error as below:

File "run_nerf.py", line 993, in train
    sigma_loss = extras_col['sigma_loss'].mean()
KeyError: 'sigma_loss'

Could you explain why not use the sigma_loss by default and give some instruction to enable it or do i miss something?

Thanks for your help.

[Beniko95J]

Hi, I also found that sigma_loss can not be used in the code of current version. It seems that depth needs to be passed into render(...) function. Have you succeeded to use sigma_loss?

[darui18]

I have the same question

[CSU-NXY]

Same question here. Can anyone share how to use sigma_loss?

[LongruiDong]

Same question here. Can anyone share how to use sigma_loss?

I think you can find here

[sethgi]

@LongruiDong the file you link computes the loss, but this line seems to make it impossible for that to ever run since ray_batch always has 11 columns:

https://github.com/dunbar12138/DSNeRF/blob/f31225b1a1150c9d8e34a74a0b0c8b30dc19ff0f/run_nerf.py#L462

[emjay73]

I have the same question

[fanwei360]

I have the same question