pissw2016

Results 5 comments of pissw2016

I have the same problem, and inspired by you guy. Mine is because that I am training with coco.h5 and I copy this weight file to another dir sothat my...

After debug with the full `class SegmentationGenerator(Sequence):` I realize that with the official VOC2012 data set. You have to not only make your own `self.image_path_list` and `self.label_path_list`. You have to...

I found out it is the problem on loss function, I make generator produce onehot label directly and then use `'categorical_crossentropy'` as loss function in compile. It is fine.

This is because your data folder is different from the author, so in utils.py Class SegmentationGenerator your image_path_list and label_path_list is empty. while the author got a different folder from...

Tesla K80 12Gwith [Golbstein generator](https://github.com/Golbstein/Keras-segmentation-deeplab-v3.1) can run with 2 batch. I am working on it for days, I tried tf1 and tf2, I also Tried with 1080Ti, and all driver...