Inheritance diagram for gcn_utils.GCNBlock:
Public Member Functions | |
| __init__ (self, in_channels, out_channels, kernel_size, t_kernel_size=1, t_stride=1, t_padding=0, t_dilation=1, bias=True) | |
| forward (self, x, A) | |
Public Attributes | |
| kernel_size | |
| conv | |
Detailed Description
The basic module for applying a graph convolution.
Args:
in_channels (int): Number of channels in the input sequence data
out_channels (int): Number of channels produced by the convolution
kernel_size (int): Size of the graph convolving kernel
t_kernel_size (int): Size of the temporal convolving kernel
t_stride (int, optional): Stride of the temporal convolution. Default: 1
t_padding (int, optional): Temporal zero-padding added to both sides of
the input. Default: 0
t_dilation (int, optional): Spacing between temporal kernel elements.
Default: 1
bias (bool, optional): If ``True``, adds a learnable bias to the output.
Default: ``True``
Shape:
- Input[0]: Input graph sequence in :math:`(N, in_channels, T_{in}, V)` format
- Input[1]: Input graph adjacency matrix in :math:`(K, V, V)` format
- Output[0]: Outpu graph sequence in :math:`(N, out_channels, T_{out}, V)` format
- Output[1]: Graph adjacency matrix for output data in :math:`(K, V, V)` format
where
:math:`N` is a batch size,
:math:`K` is the spatial kernel size, as :math:`K == kernel_size[1]`,
:math:`T_{in}/T_{out}` is a length of input/output sequence,
:math:`V` is the number of graph nodes.
Definition at line 4 of file gcn_utils.py.
Constructor & Destructor Documentation
◆ __init__()
| gcn_utils.GCNBlock.__init__ | ( | self, | |
| in_channels, | |||
| out_channels, | |||
| kernel_size, | |||
| t_kernel_size = 1, | |||
| t_stride = 1, | |||
| t_padding = 0, | |||
| t_dilation = 1, | |||
| bias = True ) |
Definition at line 34 of file gcn_utils.py.
42 bias=True):
43 super().__init__()
44
45 self.kernel_size = kernel_size
46 self.conv = nn.Conv2d(
47 in_channels,
48 out_channels * kernel_size,
49 kernel_size=(t_kernel_size, 1),
50 padding=(t_padding, 0),
51 stride=(t_stride, 1),
52 dilation=(t_dilation, 1),
53 bias=bias)
54
Member Function Documentation
◆ forward()
| gcn_utils.GCNBlock.forward | ( | self, | |
| x, | |||
| A ) |
Definition at line 55 of file gcn_utils.py.
55 def forward(self, x, A):
56 assert A.size(0) == self.kernel_size
57
58 x = self.conv(x)
59
60 n, kc, t, v = x.size()
61 x = x.view(n, self.kernel_size, kc//self.kernel_size, t, v)
62 x = torch.einsum('nkctv,kvw->nctw', (x, A))
63
64 return x.contiguous(), A
Member Data Documentation
◆ conv
| gcn_utils.GCNBlock.conv |
Definition at line 46 of file gcn_utils.py.
◆ kernel_size
| gcn_utils.GCNBlock.kernel_size |
Definition at line 45 of file gcn_utils.py.
The documentation for this class was generated from the following file:
- smaction/utils/gcn_utils.py
