torchssim.py

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

from fused_ssim import fused_ssim, fused_ssim_opt


# 计算一维的高斯分布向量
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()


# 创建高斯核，通过两个一维高斯分布向量进行矩阵乘法得到
# 可以设定channel参数拓展为3通道
def create_window(window_size, channel=1):
    _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


# 计算SSIM
# 直接使用SSIM的公式，但是在计算均值时，不是直接求像素平均值，而是采用归一化的高斯核卷积来代替。
# 在计算方差和协方差时用到了公式Var(X)=E[X^2]-E[X]^2, cov(X,Y)=E[XY]-E[X]E[Y].
# 正如前面提到的，上面求期望的操作采用高斯核卷积代替。
def ssim(img1, img2, window_size=11, window=None, size_average=True, full=False, val_range=None):
    # Value range can be different from 255. Other common ranges are 1 (sigmoid) and 2 (tanh).
    if val_range is None:
        if torch.max(img1) > 128:
            max_val = 255
        else:
            max_val = 1

        if torch.min(img1) < -0.5:
            min_val = -1
        else:
            min_val = 0
        L = max_val - min_val
    else:
        L = val_range

    padd = 0
    (_, channel, height, width) = img1.size()
    if window is None:
        real_size = min(window_size, height, width)
        window = create_window(real_size, channel=channel).to(img1.device)

    # import ipdb; ipdb.set_trace()
    mu1 = F.conv2d(img1, window, padding=padd, groups=channel)
    mu2 = F.conv2d(img2, window, padding=padd, groups=channel)

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

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

    C1 = (0.01 * L) ** 2
    C2 = (0.03 * L) ** 2

    v1 = 2.0 * sigma12 + C2
    v2 = sigma1_sq + sigma2_sq + C2
    cs = torch.mean(v1 / v2)  # contrast sensitivity

    ssim_map = ((2 * mu1_mu2 + C1) * v1) / ((mu1_sq + mu2_sq + C1) * v2)

    if size_average:
        ret = ssim_map.mean()
    else:
        ret = ssim_map.mean(1).mean(1).mean(1)

    if full:
        return ret, cs
    return ret



# Classes to re-use window
class SSIM(torch.nn.Module):
    def __init__(self, window_size=11, size_average=True, val_range=None):
        super(SSIM, self).__init__()
        self.window_size = window_size
        self.size_average = size_average
        self.val_range = val_range

        # Assume 1 channel for SSIM
        self.channel = 1
        self.window = create_window(window_size)

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

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

        return ssim(img1, img2, window=window, window_size=self.window_size, size_average=self.size_average)
    

if __name__ == "__main__":
    torch.manual_seed(0)
    # gt_image1 = torch.rand(1, 3, 1080, 1920).cuda()
    gt_image1 = torch.rand(1, 3, 2160, 3840).cuda()
    predicted_image1 = torch.nn.Parameter(torch.rand_like(gt_image1)).cuda()
    
    window_size = 11
    window = create_window(window_size, gt_image1.size(1)).to(gt_image1.device).type(gt_image1.dtype)
    ssim_value_ref1 = ssim(predicted_image1, gt_image1, window=window, window_size=window_size, size_average=True)


    ssim_value1 = fused_ssim(predicted_image1, gt_image1)
    torch.cuda.nvtx.range_push("fused_ssim")
    ssim_value2= fused_ssim_opt(predicted_image1, gt_image1)
    torch.cuda.nvtx.range_pop()
    # if ssim_value1 != ssim_value2:
    #     print("Error")

    # print(f"ref: {ssim_value_ref1}\nfused_ssim1: {ssim_value1}")
    print(f"ref: {ssim_value_ref1}\nfused_ssim1: {ssim_value1}\nfused_ssim2: {ssim_value2}")

    # """
    # Timing
    times = 1000
    start = torch.cuda.Event(enable_timing=True)
    end = torch.cuda.Event(enable_timing=True)
    start.record()
    for _ in range(times):
        fused_ssim(predicted_image1, gt_image1)
    end.record()
    torch.cuda.synchronize()
    print("fused_ssim elapsed time: ", start.elapsed_time(end)/times, "ms")

    start.record()
    for _ in range(times):
        fused_ssim_opt(predicted_image1, gt_image1)
    end.record()
    torch.cuda.synchronize()
    print("fused_ssim_opt elapsed time: ", start.elapsed_time(end)/times, "ms")
    # """