TDNN代码

核心思想是将语音的提取特征的帧进行前后联系，展开

举例：
1，2，3，4，5

123，234，345

进行了扩展，使得网络看到的特征范围更广

import torch.nn as nn
import torch.nn.function as F

class TDNN(nn.Moudle):

    def __init__(
                    self,
                    input_dim,
                    output_dim,
                    context_size,
                    stride,
                    dilation,
                    batch_norm,
                    dropout            
                ):
        super(TDNN, self).__init__()
        self.input_dim = input_dim
        self.output_dim = output_dim
        self.context_size = context_size
        self.stride = stride
        self.dilation = dilation
        self.batch_norm = batch_norm
        self.dropout = dropout

        self.kernel = nn.Linear(input_dim * context_size, output_dim)
        self.nonlinearity = nn.ReLU()
        if self.batch_norm:
            self.bn = nn.BatchNorm1d(output_dim)
        if self.dropout_p:
            self.dropout = nn.Dropout(p = self.dropout)
    
    def forward(self, x):
        _, _, d = x.shape
        x = x.unsqueeze(1)

        # 将前后几帧都联系起来，已当前帧为中心，扩展到周围
        x = F.unfold(
                    x,
                    (self.context_size, self.input_dim),
                    stride = (1, self.input_dim),
                    dilation = (self.dilation, 1)
                    ) 

        x = x.tranpose(1, 2)
        x = self.kernel(x.float())
        x = self.nonlinearity(x)

        if self.dropout_p:
            x = self.dropout(x)
        
        if self.batch_norm:
            x = x.transpose(1, 2)
            x = self.bn(x)
            x = x.transpose(1, 2)
        
        return x

关于unfold函数的理解

http://t.csdnimg.cn/HAe40

http://t.csdnimg.cn/2wuke

关于transpose函数的理解

http://t.csdnimg.cn/C8Xd6