Does your test loss decline regularly? about mvsnet_pytorch HOT 5 CLOSED

I have re-train your model using the same setting as yours, and I test it for evaluating the depth map error on test dataset.
I find out your depth map error is from 8 to 9 using your provided model or re-trained model, but the original MVSNet depth map error is about 4 to 5.
Through the visualization of tensorboard, it seems that your model does not train well.

from mvsnet_pytorch.

Sorry that I don't have much time to verify your problem. Maybe because of the pytorch verison? I used 0.4.1 for this code.

from mvsnet_pytorch.

@yhw-yhw Hello, I have the same problem as you. Have you solved this problem? The testing loss is confusing.

from mvsnet_pytorch.

Cause the differences between this validation code and the original code, I remember that one is depth range constraint, the other is the different dataset's number. You can compare this validation code carefully with the original validation code.

from mvsnet_pytorch.

@yhw-yhw what's the validation code and the original version?

from mvsnet_pytorch.