loss.py

import torch
from torch import nn
import numpy as np
import random
import torch.nn.functional as F
from math import exp


def _iou(pred, target, size_average=True):
    b = pred.shape[0]
    IoU = 0.0
    for i in range(0, b):
        # compute the IoU of the foreground
        Iand1 = torch.sum(target[i, :, :, :] * pred[i, :, :, :])
        Ior1 = torch.sum(target[i, :, :, :]) + torch.sum(pred[i, :, :, :]) - Iand1
        IoU1 = Iand1 / Ior1

        # IoU loss is (1-IoU1)
        IoU = IoU + (1 - IoU1)

    return IoU / b


class IOU(torch.nn.Module):
    def __init__(self, size_average=True):
        super(IOU, self).__init__()
        self.size_average = size_average

    def forward(self, pred, target):
        return _iou(pred, target, self.size_average)


def gaussian(window_size, sigma):
    gauss = torch.Tensor([exp(-(x - window_size // 2) ** 2 / float(2 * sigma ** 2)) for x in range(window_size)])
    return gauss / gauss.sum()


def create_window(window_size, channel):
    _1D_window = gaussian(window_size, 1.5).unsqueeze(1)
    _2D_window = _1D_window.mm(_1D_window.t()).float().unsqueeze(0).unsqueeze(0)
    window = _2D_window.expand(channel, 1, window_size, window_size).contiguous()
    return window


def _ssim(img1, img2, window, window_size, channel, size_average=True):
    mu1 = F.conv2d(img1, window, padding=window_size // 2, groups=channel)
    mu2 = F.conv2d(img2, window, padding=window_size // 2, groups=channel)

    mu1_sq = mu1.pow(2)
    mu2_sq = mu2.pow(2)
    mu1_mu2 = mu1 * mu2

    sigma1_sq = F.conv2d(img1 * img1, window, padding=window_size // 2, groups=channel) - mu1_sq
    sigma2_sq = F.conv2d(img2 * img2, window, padding=window_size // 2, groups=channel) - mu2_sq
    sigma12 = F.conv2d(img1 * img2, window, padding=window_size // 2, groups=channel) - mu1_mu2

    C1 = 0.01 ** 2
    C2 = 0.03 ** 2

    ssim_map = ((2 * mu1_mu2 + C1) * (2 * sigma12 + C2)) / ((mu1_sq + mu2_sq + C1) * (sigma1_sq + sigma2_sq + C2))

    if size_average:
        return ssim_map.mean()
    else:
        return ssim_map.mean(1).mean(1).mean(1)


class SSIM(torch.nn.Module):
    def __init__(self, window_size=11, size_average=True):
        super(SSIM, self).__init__()
        self.window_size = window_size
        self.size_average = size_average
        self.channel = 1
        self.window = create_window(window_size, self.channel)

    def forward(self, img1, img2):
        (_, channel, _, _) = img1.size()

        if channel == self.channel and self.window.data.type() == img1.data.type():
            window = self.window
        else:
            window = create_window(self.window_size, channel)

            if img1.is_cuda:
                window = window.cuda(img1.get_device())
            window = window.type_as(img1)

            self.window = window
            self.channel = channel

        return _ssim(img1, img2, window, self.window_size, channel, self.size_average)


def bce_iou_loss(pred, target):
    bce_loss = nn.BCELoss(size_average=True)
    iou_loss_f = IOU(size_average=True)
    bce_out = bce_loss(pred, target)
    iou_out = iou_loss_f(pred, target)

    loss = bce_out + iou_out

    return loss


def bce_ssim_loss(pred, target):
    bce_loss = nn.BCELoss(size_average=True)
    ssim_loss = SSIM(window_size=11, size_average=True)
    iou_loss = IOU(size_average=True)
    bce_out = bce_loss(pred, target)
    ssim_out = 1 - ssim_loss(pred, target)
    iou_out = iou_loss(pred, target)
    loss = bce_out + iou_out + ssim_out
    return loss


def structure_loss(pred, mask):
    weit = 1 + 5 * torch.abs(F.avg_pool2d(mask, kernel_size=31, stride=1, padding=15) - mask)
    wbce = F.binary_cross_entropy(pred, mask, reduce='none')
    wbce = (weit * wbce).sum(dim=(2, 3)) / weit.sum(dim=(2, 3))

    inter = ((pred * mask) * weit).sum(dim=(2, 3))
    union = ((pred + mask) * weit).sum(dim=(2, 3))
    wiou = 1 - (inter + 1) / (union - inter + 1)
    return (wbce + wiou).mean()


def build_loss(loss_type='bi'):
    if loss_type == 'bce':
        return nn.BCELoss(size_average=True)
    elif loss_type == 'bi':
        return bce_iou_loss
    elif loss_type == 'bas':
        return bce_ssim_loss
    elif loss_type == 'f3':
        return structure_loss