face-parsing.PyTorch
face-parsing.PyTorch copied to clipboard
The class number is 18, why set `BiSeNet(19)`
明明只有18个类别,为啥train.py和test.py都设置成下面这样:
n_classes = 19
net = BiSeNet(n_classes=n_classes)
另外为啥把HQ数据集的background标签去掉了,加上不是正好19个嘛??
@piwawa , in segmentation you need to add background also as a class.