some error use connectivity loss fuction in gpu about indexnet_matting HOT 12 OPEN

Hi @kelisiya,
Sorry for the late reply due to the CVPR deadline.
Do you mean the training loss? or calculating the evaluation errors?
The connectivity loss is not used in training, it is only used when evaluating the matte quality. It should be fine if following my instructions to run the code.

from indexnet_matting.

Hi @kelisiya,
Sorry for the late reply due to the CVPR deadline.
Do you mean the training loss? or calculating the evaluation errors?
The connectivity loss is not used in training, it is only used when evaluating the matte quality. It should be fine if following my instructions to run the code.

It's work , I let your backbone to DIM and use Alpha loss and Grad loss , finally I can train your model .

from indexnet_matting.

@kelisiya Nice, let me know if your model achieves better results:)

from indexnet_matting.

from indexnet_matting.

@kelisiya It is normal that the network's output is not bounded by [0, 1] due to the nature of regression. This is why postprocessing is required to eliminate unreasonale outputs.
However, you should not use the clip operator or clamp in pytorch during training, because the gradient will be clipped to zero either. This may be the reason why the loss does not decrease.
The clip operator should be only applicable in inference.

from indexnet_matting.

from indexnet_matting.

I don't use cv.normalize.
I also tried sigmoid, but do not find it necessary.

from indexnet_matting.

from indexnet_matting.

Exactly!

from indexnet_matting.

from indexnet_matting.

from indexnet_matting.

Of course you should normalize the alpha before calculating the loss.
Yes, the concatenated input like DIM.

from indexnet_matting.