model_convlstm.py

import torch.nn as nn
import torch

init_func = {
    'uniform':nn.init.uniform_,
    'normal':nn.init.normal_,
    'constant':nn.init.constant_,
    'xavier_uniform': nn.init.xavier_uniform_,
    'xavier_normal': nn.init.xavier_normal_,
    'kaiming_uniform': nn.init.kaiming_uniform_,
    'kaiming_normal': nn.init.kaiming_normal_,
    'orthogonal': nn.init.orthogonal_,
}

def init_weights(model, funcname='xavier_uniform'):
    for m in model.modules():
        if isinstance(m, nn.Conv2d):
            init_func[funcname](m.weight)
            if m.bias is not None:
                nn.init.constant_(m.bias, 0.)
        elif isinstance(m, nn.BatchNorm2d):
            nn.init.constant_(m.weight, 1.)
            nn.init.constant_(m.bias, 0.)
        elif isinstance(m, nn.Linear):
            nn.init.nn.init.normal_(m.weight, 0, 0.01)
            nn.init.constant_(m.bias, 0.)


class ConvLSTMCell(nn.Module):

    def __init__(self, input_size, input_dim, hidden_dim, kernel_size, bias):
        """
        Initialize ConvLSTM cell.

        Parameters
        ----------
        input_size: (int, int)
            Height and width of input tensor as (height, width).
        input_dim: int
            Number of channels of input tensor.
        hidden_dim: int
            Number of channels of hidden state.
        kernel_size: (int, int)
            Size of the convolutional kernel.
        bias: bool
            Whether or not to add the bias.
        """

        super(ConvLSTMCell, self).__init__()

        self.height, self.width = input_size
        self.input_dim  = input_dim
        self.hidden_dim = hidden_dim

        self.kernel_size = kernel_size
        self.padding     = kernel_size[0] // 2, kernel_size[1] // 2
        self.bias        = bias

        self.rnn_conv = nn.Conv2d(in_channels=self.input_dim + self.hidden_dim,
                              out_channels=4 * self.hidden_dim,
                              kernel_size=self.kernel_size,
                              padding=self.padding,
                              bias=self.bias)

        init_weights(self.rnn_conv, 'xavier_uniform')

    def forward(self, input_tensor, cur_state):

        h_cur, c_cur = cur_state
        combined = torch.cat([input_tensor, h_cur], dim=1)  # concatenate along channel axis

        combined_conv = self.rnn_conv(combined)
        cc_i, cc_f, cc_o, cc_g = torch.split(combined_conv, self.hidden_dim, dim=1)
        i = torch.sigmoid(cc_i)
        f = torch.sigmoid(cc_f)
        o = torch.sigmoid(cc_o)
        g = torch.tanh(cc_g)

        c_next = f * c_cur + i * g
        h_next = o * torch.tanh(c_next)

        return h_next, c_next

    def init_hidden(self, batch_size):
        return (torch.zeros(batch_size, self.hidden_dim, self.height, self.width).cuda(),
                torch.zeros(batch_size, self.hidden_dim, self.height, self.width).cuda())

class ConvLSTM(nn.Module):

    def __init__(self, input_size, input_dim, hidden_dim, kernel_size, num_layers,
                 batch_first=False, bias=True, return_all_layers=False):
        super(ConvLSTM, self).__init__()

        self._check_kernel_size_consistency(kernel_size)

        # Make sure that both `kernel_size` and `hidden_dim` are lists having len == num_layers
        kernel_size = self._extend_for_multilayer(kernel_size, num_layers)
        hidden_dim = self._extend_for_multilayer(hidden_dim, num_layers)
        if not len(kernel_size) == len(hidden_dim) == num_layers:
            raise ValueError('Inconsistent list length.')

        self.height, self.width = input_size

        self.input_dim = input_dim
        self.hidden_dim = hidden_dim
        self.kernel_size = kernel_size
        self.num_layers = num_layers
        self.batch_first = batch_first
        self.bias = bias
        self.return_all_layers = return_all_layers

        cell_list = []
        for i in range(0, self.num_layers):
            cur_input_dim = self.input_dim if i == 0 else self.hidden_dim[i - 1]

            cell_list.append(ConvLSTMCell(input_size=(self.height, self.width),
                                          input_dim=cur_input_dim,
                                          hidden_dim=self.hidden_dim[i],
                                          kernel_size=self.kernel_size[i],
                                          bias=self.bias))

        self.cell_list = nn.ModuleList(cell_list)

    def forward(self, input_tensor, hidden_state=None):
        """

        Parameters
        ----------
        input_tensor: todo
            5-D Tensor either of shape (t, b, c, h, w) or (b, t, c, h, w)
        hidden_state: todo
            None. todo implement stateful

        Returns
        -------
        last_state_list, layer_output
        """
        if not self.batch_first:
            # (t, b, c, h, w) -> (b, t, c, h, w)
            input_tensor = input_tensor.permute(1, 0, 2, 3, 4)

        # Implement stateful ConvLSTM
        if hidden_state is not None:
            pass
            # raise NotImplementedError()
        else:
            hidden_state = self._init_hidden(batch_size=input_tensor.size(0))

        layer_output_list = []
        last_state_list = []

        seq_len = input_tensor.size(1)
        cur_layer_input = input_tensor

        for layer_idx in range(self.num_layers):

            h, c = hidden_state[layer_idx]
            output_inner = []
            for t in range(seq_len):
                h, c = self.cell_list[layer_idx](input_tensor=cur_layer_input[:, t, :, :, :],
                                                 cur_state=[h, c])
                output_inner.append(h)

            layer_output = torch.stack(output_inner, dim=1)
            cur_layer_input = layer_output

            layer_output_list.append(layer_output)
            last_state_list.append([h, c])

        if not self.return_all_layers:
            layer_output_list = layer_output_list[-1]
            last_state_list = last_state_list[-1]

        return layer_output_list, last_state_list

    def _init_hidden(self, batch_size):
        init_states = []
        for i in range(self.num_layers):
            init_states.append(self.cell_list[i].init_hidden(batch_size))
        return init_states

    @staticmethod
    def _check_kernel_size_consistency(kernel_size):
        if not (isinstance(kernel_size, tuple) or
                (isinstance(kernel_size, list) and all([isinstance(elem, tuple) for elem in kernel_size]))):
            raise ValueError('`kernel_size` must be tuple or list of tuples')

    @staticmethod
    def _extend_for_multilayer(param, num_layers):
        if not isinstance(param, list):
            param = [param] * num_layers
        return param

class AttConvLSTMCell(nn.Module):

    def __init__(self, input_size, input_dim, hidden_dim, att_dim, kernel_size, bias):
        """
        Initialize ConvLSTM cell.

        Parameters
        ----------
        input_size: (int, int)
            Height and width of input tensor as (height, width).
        input_dim: int
            Number of channels of input tensor.
        hidden_dim: int
            Number of channels of hidden state.
        kernel_size: (int, int)
            Size of the convolutional kernel.
        bias: bool
            Whether or not to add the bias.
        """

        super(AttConvLSTMCell, self).__init__()

        self.height, self.width = input_size
        self.input_dim = input_dim
        self.hidden_dim = hidden_dim
        self.att_dim = att_dim

        self.kernel_size = kernel_size
        self.padding = kernel_size[0] // 2, kernel_size[1] // 2
        self.bias = bias

        self.rnn_conv = nn.Conv2d(in_channels=self.input_dim + self.hidden_dim,
                              out_channels=4 * self.hidden_dim,
                              kernel_size=self.kernel_size,
                              padding=self.padding,
                              bias=self.bias)

        self.att_conv = nn.Conv2d(in_channels=self.input_dim + self.hidden_dim,
                              out_channels=self.att_dim,
                              kernel_size=self.kernel_size,
                              padding=self.padding,
                              bias=self.bias)

        self.att_conv_out = nn.Conv2d(self.att_dim, 1, kernel_size=self.kernel_size, padding=self.padding,bias=False)

        init_weights(self.rnn_conv, 'xavier_uniform')
        init_weights(self.att_conv, 'xavier_uniform')
        init_weights(self.att_conv_out,'xavier_uniform')


    def forward(self, input_tensor, cur_state):
        h_cur, c_cur = cur_state

        # For attention block
        att_map = self.att_conv_out(torch.tanh(self.att_conv(torch.cat([input_tensor, h_cur], dim=1))))
        batch, channel, height, width = att_map.size()
        att_map = torch.softmax(att_map.view(batch,1,height*width),2)
        att_map = att_map.view(batch, 1, height, width)


        att_input = input_tensor * att_map.repeat(1,input_tensor.size(1),1,1)
        combined = torch.cat([att_input, h_cur], dim=1)  # concatenate along channel axis

        combined_conv = self.rnn_conv(combined)
        cc_i, cc_f, cc_o, cc_g = torch.split(combined_conv, self.hidden_dim, dim=1)
        i = torch.sigmoid(cc_i)
        f = torch.sigmoid(cc_f)
        o = torch.sigmoid(cc_o)
        g = torch.tanh(cc_g)

        c_next = f * c_cur + i * g
        h_next = o * torch.tanh(c_next)

        return h_next, c_next

    def init_hidden(self, batch_size):
        return (torch.zeros(batch_size, self.hidden_dim, self.height, self.width).cuda(),
                torch.zeros(batch_size, self.hidden_dim, self.height, self.width).cuda())

class AttConvLSTM(nn.Module):

    def __init__(self, input_size, input_dim, hidden_dim, att_dim, kernel_size, num_layers,
                 batch_first=False, bias=True, return_all_layers=False):
        super(AttConvLSTM, self).__init__()

        self._check_kernel_size_consistency(kernel_size)

        # Make sure that both `kernel_size` and `hidden_dim` are lists having len == num_layers
        kernel_size = self._extend_for_multilayer(kernel_size, num_layers)
        hidden_dim = self._extend_for_multilayer(hidden_dim, num_layers)
        if not len(kernel_size) == len(hidden_dim) == num_layers:
            raise ValueError('Inconsistent list length.')

        self.height, self.width = input_size

        self.input_dim = input_dim
        self.hidden_dim = hidden_dim
        self.att_dim = att_dim
        self.kernel_size = kernel_size
        self.num_layers = num_layers
        self.batch_first = batch_first
        self.bias = bias
        self.return_all_layers = return_all_layers

        cell_list = []
        for i in range(0, self.num_layers):
            cur_input_dim = self.input_dim if i == 0 else self.hidden_dim[i - 1]

            cell_list.append(AttConvLSTMCell(input_size=(self.height, self.width),
                                             input_dim=cur_input_dim,
                                             hidden_dim=self.hidden_dim[i],
                                             att_dim=att_dim,
                                             kernel_size=self.kernel_size[i],
                                             bias=self.bias))

        self.cell_list = nn.ModuleList(cell_list)

    def forward(self, input_tensor, hidden_state=None):
        """

        Parameters
        ----------
        input_tensor: todo
            5-D Tensor either of shape (t, b, c, h, w) or (b, t, c, h, w)
        hidden_state: todo
            None. todo implement stateful

        Returns
        -------
        last_state_list, layer_output
        """
        if not self.batch_first:
            # (t, b, c, h, w) -> (b, t, c, h, w)
            input_tensor = input_tensor.permute(1, 0, 2, 3, 4)

        # Implement stateful ConvLSTM
        if hidden_state is not None:
            pass
            # raise NotImplementedError()
        else:
            hidden_state = self._init_hidden(batch_size=input_tensor.size(0))

        layer_output_list = []
        last_state_list = []

        seq_len = input_tensor.size(1)
        cur_layer_input = input_tensor

        for layer_idx in range(self.num_layers):

            h, c = hidden_state[layer_idx]
            output_inner = []
            for t in range(seq_len):
                h, c = self.cell_list[layer_idx](input_tensor=cur_layer_input[:, t, :, :, :], cur_state=[h, c])
                output_inner.append(h)

            layer_output = torch.stack(output_inner, dim=1)
            cur_layer_input = layer_output

            layer_output_list.append(layer_output)
            last_state_list.append([h, c])

        if not self.return_all_layers:
            layer_output_list = layer_output_list[-1]
            last_state_list = last_state_list[-1]

        return layer_output_list, last_state_list

    def _init_hidden(self, batch_size):
        init_states = []
        for i in range(self.num_layers):
            init_states.append(self.cell_list[i].init_hidden(batch_size))
        return init_states

    @staticmethod
    def _check_kernel_size_consistency(kernel_size):
        if not (isinstance(kernel_size, tuple) or
                (isinstance(kernel_size, list) and all([isinstance(elem, tuple) for elem in kernel_size]))):
            raise ValueError('`kernel_size` must be tuple or list of tuples')

    @staticmethod
    def _extend_for_multilayer(param, num_layers):
        if not isinstance(param, list):
            param = [param] * num_layers
        return param