Dataset Setting

[Hardcandies]

Hi,

In voc_tarin.py, you filter out the image that include noval categories using function 'filte_multi_class'. Is that right?

What is the reason for filtering? And what are the effects if you don't filter out some images?

Can other paper such as CANet also use this setting?

jachymchen

[Yang-Bob]

Hi,

For few-shot semantic segmentation, all categories in the dataset will be divided as base classes or novel classes. The base classes are used to train network and novel classes which have few labeled images are used to test. Actually, some images have multi-class objects. We do this operation in order to select the images which only have base class objects to train the network.

We followed this setting from SG-One. And I don't think there is any problem with it under the few-shot semantic segmentation task.

[Hardcandies]

OK. Thanks for your reply.

[Yang-Bob]

My pleasure.