Section Navigation
Global Second-order Pooling Convolutional Networks from [Gao2018].
in_channels (int,) – number of input feature channels
reduction_rate (int,) – reduction ratio of the fully-connected layers
bias (bool, default=False) – if True, adds a learnable bias will be used in the convolution.
References
Gao, Z., Jiangtao, X., Wang, Q., Li, P., 2018.
Global Second-order Pooling Convolutional Networks. CVPR 2018.
Methods
Apply the Global Second-order Pooling Convolutional Networks block.
x (Pytorch.Tensor)
Pytorch.Tensor
GSoP
GSoP.forward()