Own dataset training produces low loss, but unsatisfying results about icnet-tensorflow HOT 5 CLOSED

Hey @Tamme, where did these images come from, output results generated by the model? Because it seems that the first one is much better than the second one. So your dataset is compressed version of cityscapes dataset?

from icnet-tensorflow.

Sry I didnt mention. It is a forest dataset. The first is ground truth and second is output of icnet. and the datasets classes are a subset of cityscapes with 6 classes.

from icnet-tensorflow.

I got it, I think the problem just as you mentioned above. You can train from the very beginning without loading pre-trained weights, since two dataset are very different.

from icnet-tensorflow.

@Tamme can you elaborate steps you did to update the number of classes so training works? Also did u ever got over the inference and loss contradiction issue you faced and mentioned above?

from icnet-tensorflow.

@ogail, to adjust the number of classes, I modified the load function in network.py. Specifically, if the op name is 'conv6_cls' then I picked the classes the I desired.

The contradiction issue was just a bug I had in the training data. I guess you just have to go over everything and see if all makes sense.

from icnet-tensorflow.