# Copyright 2021-2022 The Alibaba Fundamental Vision Team Authors. All rights reserved.
import cv2
import numpy as np
import torch.nn as nn
import torch.nn.functional as F


def gn_init(m, zero_init=False):
    assert isinstance(m, nn.GroupNorm)
    m.weight.data.fill_(0. if zero_init else 1.)
    m.bias.data.zero_()


class Bottleneck(nn.Module):
    expansion = 4

    def __init__(self, inplanes, planes, stride=1, downsample=None):
        """Bottleneck for resnet-style networks
        Args:
            inplanes: input channel number
            planes: output channel number
        """
        super(Bottleneck, self).__init__()
        self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, bias=False)
        self.bn1 = nn.GroupNorm(32, planes)
        self.conv2 = nn.Conv2d(
            planes,
            planes,
            kernel_size=3,
            stride=stride,
            padding=1,
            bias=False)
        self.bn2 = nn.GroupNorm(32, planes)
        self.conv3 = nn.Conv2d(planes, planes * 4, kernel_size=1, bias=False)
        self.bn3 = nn.GroupNorm(32, planes * 4)
        self.relu = nn.ReLU(inplace=True)
        self.downsample = downsample
        self.stride = stride

        gn_init(self.bn1)
        gn_init(self.bn2)
        gn_init(self.bn3, zero_init=True)

    def forward(self, x):
        residual = x

        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)
        out = self.relu(out)

        out = self.conv3(out)
        out = self.bn3(out)

        if self.downsample is not None:
            residual = self.downsample(x)

        out += residual
        out = self.relu(out)

        return out


class ResNet(nn.Module):
    """
    resnet-style network with group normalization
    """

    def __init__(self, block, layers, num_classes=1000):
        self.inplanes = 64
        super(ResNet, self).__init__()
        self.conv1 = nn.Conv2d(
            3, 64, kernel_size=7, stride=2, padding=3, bias=False)
        self.bn1 = nn.GroupNorm(32, 64)
        self.relu = nn.ReLU(inplace=True)
        self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
        self.layer1 = self._make_layer(block, 64, layers[0])
        self.layer2 = self._make_layer(block, 128, layers[1], stride=2)
        self.layer3 = self._make_layer(block, 256, layers[2], stride=2)
        self.layer4 = self._make_layer(block, 512, layers[3], stride=1)

        self.gap = nn.AvgPool2d((14, 14))
        self.reduce_conv = nn.Conv2d(2048, 512, kernel_size=1)

        gn_init(self.bn1)

    def _make_layer(self, block, planes, blocks, stride=1):
        downsample = None
        if stride != 1 or self.inplanes != planes * block.expansion:
            downsample = nn.Sequential(
                nn.AvgPool2d(stride, stride),
                nn.Conv2d(
                    self.inplanes,
                    planes * block.expansion,
                    kernel_size=1,
                    stride=1,
                    bias=False),
                nn.GroupNorm(32, planes * block.expansion),
            )

        layers = []
        layers.append(block(self.inplanes, planes, stride, downsample))
        self.inplanes = planes * block.expansion
        for i in range(1, blocks):
            layers.append(block(self.inplanes, planes))

        return nn.Sequential(*layers)

    def forward(self, x):
        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)
        x = self.maxpool(x)

        x = self.layer1(x)
        x = self.layer2(x)
        x = self.layer3(x)
        x = self.layer4(x)

        x = self.gap(x)
        x = self.reduce_conv(x)  # 512

        x = x.view(x.size(0), -1)  # 512
        return F.normalize(x, p=2, dim=1)


def preprocess(img):
    """
    preprocess the image with cv2-bgr style to tensor
    """
    mean = np.array([0.485, 0.456, 0.406])
    std = np.array([0.229, 0.224, 0.225])

    img_size = 224
    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
    img_new = cv2.resize(
        img, (img_size, img_size), interpolation=cv2.INTER_LINEAR)
    content = np.array(img_new).astype(np.float32)
    content = (content / 255.0 - mean) / std
    # transpose
    img_new = content.transpose(2, 0, 1)
    img_new = img_new[np.newaxis, :, :, :]
    return img_new


def resnet50_embed():
    """
    create resnet50 network with group normalization
    """
    net = ResNet(Bottleneck, [3, 4, 6, 3])
    return net