Yolov5 utils/activations.py解析

与激活函数实现相关代码

对于激活函数得重新实现，防止在有些模型中无法直接调用nn自带函数

# Activation functions

import torch
import torch.nn as nn
import torch.nn.functional as F


# SiLU https://arxiv.org/pdf/1606.08415.pdf ----------------------------------------------------------------------------
class SiLU(nn.Module):  # export-friendly version of nn.SiLU()
    @staticmethod
    def forward(x):
        return x * torch.sigmoid(x)


class Hardswish(nn.Module):  # export-friendly version of nn.Hardswish()
    @staticmethod
    def forward(x):
        # return x * F.hardsigmoid(x)  # for torchscript and CoreML
        return x * F.hardtanh(x + 3, 0., 6.) / 6.  # for torchscript, CoreML and ONNX


class MemoryEfficientSwish(nn.Module):
    class F(torch.autograd.Function):
        @staticmethod
        def forward(ctx, x):
            ctx.save_for_backward(x)
            return x * torch.sigmoid(x)

        @staticmethod
        def backward(ctx, grad_output):
            x = ctx.saved_tensors[0]
            sx = torch.sigmoid(x)
            return grad_output * (sx * (1 + x * (1 - sx)))

    def forward(self, x):
        return self.F.apply(x)


# Mish https://github.com/digantamisra98/Mish --------------------------------------------------------------------------
class Mish(nn.Module):
    @staticmethod
    def forward(x):
        return x * F.softplus(x).tanh()


class MemoryEfficientMish(nn.Module):
    class F(torch.autograd.Function):
        @staticmethod
        def forward(ctx, x):
            ctx.save_for_backward(x)
            return x.mul(torch.tanh(F.softplus(x)))  # x * tanh(ln(1 + exp(x)))

        @staticmethod
        def backward(ctx, grad_output):
            x = ctx.saved_tensors[0]
            sx = torch.sigmoid(x)
            fx = F.softplus(x).tanh()
            return grad_output * (fx + x * sx * (1 - fx * fx))

    def forward(self, x):
        return self.F.apply(x)


# FReLU https://arxiv.org/abs/2007.11824 -------------------------------------------------------------------------------
class FReLU(nn.Module):
    def __init__(self, c1, k=3):  # ch_in, kernel
        super().__init__()
        self.conv = nn.Conv2d(c1, c1, k, 1, 1, groups=c1, bias=False)
        self.bn = nn.BatchNorm2d(c1)

    def forward(self, x):
        return torch.max(x, self.bn(self.conv(x)))

staticmethod用于修饰类中的方法,使其可以在不创建类实例的情况下调用方法，这样做的好处是执行效率比较高。当然，也可以像一般的方法一样用实例调用该方法。该方法一般被称为静态方法。静态方法不可以引用类中的属性或方法，其参数列表也不需要约定的默认参数self。

staticmethod

Swish

class SiLU(nn.Module):  # export-friendly version of nn.SiLU()
    @staticmethod
    def forward(x):
        return x * torch.sigmoid(x)

Handswish

class Hardswish(nn.Module):  # export-friendly version of nn.Hardswish()
    @staticmethod
    def forward(x):
        # return x * F.hardsigmoid(x)  # for torchscript and CoreML
        return x * F.hardtanh(x + 3, 0., 6.) / 6.  

Mish

Mish(x)=x∗Tanh(Softplus(x))

class Mish(nn.Module):
    @staticmethod
    def forward(x):
        return x * F.softplus(x).tanh()

FReLU

三种激活函数对比

FReLU：x与T(x)求max 多了一个funnel condition的操作

class FReLU(nn.Module):
    def __init__(self, c1, k=3):  # ch_in, kernel
        super().__init__()
        # nn.Con2d(in_channels, out_channels, kernel_size, stride, padding, dilation=1, groups=1, bias=True)
        self.conv = nn.Conv2d(c1, c1, k, 1, 1, groups=c1) 
        self.bn = nn.BatchNorm2d(c1)

    def forward(self, x):
        return torch.max(x, self.bn(self.conv(x)))