Adding more classes

[trocolit2]

Dear @tersekmatija

Thanks for sharing your work and code. We have tested it with different images and situations and seems robust, giving us the expected results.

I would like to extend the number of classes, to include the boat and buoy.
Is it possible to include these 2 new classes? Which approach do you suggest to follow?

Regards

[tersekmatija]

@trocolit2

Thanks and apologies for a late reply. I somehow missed this issue in my email folder... This is certainly possible.

First, you would have to modify the dataset or provide your own, which has boat and buoy annotated. Alternatively, you can use LaRS (https://lojzezust.github.io/lars-dataset/) dataset. Note that eWaSR is semantic segmentation network, while LaRS provides instance segmentation for buoy and boat. You would have to do some dataset preprocessing to create semantic segmentation masks from instances.

Depending on your dataset format, you might need to support additional data loader here or simply adjust one of the existing ones to make it work for your case.

You should be then able to start the training with --num_classes 5 flag.