Yang-Bob / PMMs

Prototype Mixture Models for Few-shot Semantic Segmentation
163 stars 27 forks source link

question about FPMMs.py #3

Closed cymqqqq closed 4 years ago

cymqqqq commented 4 years ago
        nn.Conv2d(256+2,256,kernel_size=3,stride=1,padding=1,bias=True),

why the in_channel = 256+2?

Yang-Bob commented 4 years ago

256 means the feature channel and 2 denotes the foreground & background probability maps.

cymqqqq commented 4 years ago

thx