Fisher Yu

I am working on it. sorry about the interruption.

Did you use my caffe for training? https://github.com/fyu/caffe-dilation It is not necessary to run prediction.

@vigorous0503 if caffe_pb2 doesn't exist, you can try to make clean and build the binary again. See if that solves your problem. If you copy caffe_pb2 from original caffe, you...

@Dandue Please check README of caffe-dilation. Or https://github.com/fyu/dilation/issues/11

@Timo-hab @chenqifeng22 test_batch has to be 0 for cityscape joint training. Was it set correctly?

@chenqifeng22 You have to check your setup. A lot of people have confirmed to me that predict.py can work well.

@qualia0000 Besides the batch size, the memory consumption also depends on the crop size.

The loss of initial stages should be around 3.0 for a 19 category classification problem. If you observe something much bigger than that, it probably indicates the optimization diverges. It...

I just added an option to set iter_size in the training options: https://github.com/fyu/dilation/blob/master/train.py#L233. If your GPU doesn't have enough memory and you have to decrease the batch size, you can...

one channel