Partial Convolution (PConv卷积)

Partial Convolution (PConv卷积)，有助于提升模型对小目标检测的性能。目前许多研究都集中在减少浮点运算（FLOPs）的数量上。然而FLOPs的这种减少不一定会带来延迟的类似程度的减少。这主要源于每秒低浮点运算（FLOPS）效率低下。为了实现更快的网络，作者重新回顾了FLOPs的运算符，并证明了如此低的FLOPS主要是由于运算符的频繁内存访问，尤其是深度卷积。因此提出了一种新的partial convolution（PConv卷积），通过同时减少冗余计算和内存访问可以更有效地提取空间特征。

原文地址：Run, Don’t Walk: Chasing Higher FLOPS for Faster Neural Networks

代码实现：

class PConv(nn.Module):
    def __init__(self, dim, ouc, n_div=4, forward='split_cat'):
        super().__init__()
        self.dim_conv3 = dim // n_div
        self.dim_untouched = dim - self.dim_conv3
        self.partial_conv3 = nn.Conv2d(self.dim_conv3, self.dim_conv3, 3, 1, 1, bias=False)
        self.conv = Conv(dim, ouc, k=1)

        if forward == 'slicing':
            self.forward = self.forward_slicing
        elif forward == 'split_cat':
            self.forward = self.forward_split_cat
        else:
            raise NotImplementedError

    def forward_slicing(self, x):
        # only for inference
        x = x.clone()   # !!! Keep the original input intact for the residual connection later
        x[:, :self.dim_conv3, :, :] = self.partial_conv3(x[:, :self.dim_conv3, :, :])
        x = self.conv(x)
        return x

    def forward_split_cat(self, x):
        # for training/inference
        x1, x2 = torch.split(x, [self.dim_conv3, self.dim_untouched], dim=1)
        x1 = self.partial_conv3(x1)
        x = torch.cat((x1, x2), 1)
        x = self.conv(x)
        return x