annotator/lineart_anime/__init__.py

import numpy as np
import torch
import torch.nn as nn
import functools

import os
import cv2
from einops import rearrange
from modules import devices
from annotator.annotator_path import models_path


class UnetGenerator(nn.Module):
    """Create a Unet-based generator"""

    def __init__(self, input_nc, output_nc, num_downs, ngf=64, norm_layer=nn.BatchNorm2d, use_dropout=False):
        """Construct a Unet generator
        Parameters:
            input_nc (int)  -- the number of channels in input images
            output_nc (int) -- the number of channels in output images
            num_downs (int) -- the number of downsamplings in UNet. For example, # if |num_downs| == 7,
                                image of size 128x128 will become of size 1x1 # at the bottleneck
            ngf (int)       -- the number of filters in the last conv layer
            norm_layer      -- normalization layer
        We construct the U-Net from the innermost layer to the outermost layer.
        It is a recursive process.
        """
        super(UnetGenerator, self).__init__()
        # construct unet structure
        unet_block = UnetSkipConnectionBlock(ngf * 8, ngf * 8, input_nc=None, submodule=None, norm_layer=norm_layer, innermost=True)  # add the innermost layer
        for _ in range(num_downs - 5):          # add intermediate layers with ngf * 8 filters
            unet_block = UnetSkipConnectionBlock(ngf * 8, ngf * 8, input_nc=None, submodule=unet_block, norm_layer=norm_layer, use_dropout=use_dropout)
        # gradually reduce the number of filters from ngf * 8 to ngf
        unet_block = UnetSkipConnectionBlock(ngf * 4, ngf * 8, input_nc=None, submodule=unet_block, norm_layer=norm_layer)
        unet_block = UnetSkipConnectionBlock(ngf * 2, ngf * 4, input_nc=None, submodule=unet_block, norm_layer=norm_layer)
        unet_block = UnetSkipConnectionBlock(ngf, ngf * 2, input_nc=None, submodule=unet_block, norm_layer=norm_layer)
        self.model = UnetSkipConnectionBlock(output_nc, ngf, input_nc=input_nc, submodule=unet_block, outermost=True, norm_layer=norm_layer)  # add the outermost layer

    def forward(self, input):
        """Standard forward"""
        return self.model(input)


class UnetSkipConnectionBlock(nn.Module):
    """Defines the Unet submodule with skip connection.
        X -------------------identity----------------------
        |-- downsampling -- |submodule| -- upsampling --|
    """

    def __init__(self, outer_nc, inner_nc, input_nc=None,
                 submodule=None, outermost=False, innermost=False, norm_layer=nn.BatchNorm2d, use_dropout=False):
        """Construct a Unet submodule with skip connections.
        Parameters:
            outer_nc (int) -- the number of filters in the outer conv layer
            inner_nc (int) -- the number of filters in the inner conv layer
            input_nc (int) -- the number of channels in input images/features
            submodule (UnetSkipConnectionBlock) -- previously defined submodules
            outermost (bool)    -- if this module is the outermost module
            innermost (bool)    -- if this module is the innermost module
            norm_layer          -- normalization layer
            use_dropout (bool)  -- if use dropout layers.
        """
        super(UnetSkipConnectionBlock, self).__init__()
        self.outermost = outermost
        if type(norm_layer) == functools.partial:
            use_bias = norm_layer.func == nn.InstanceNorm2d
        else:
            use_bias = norm_layer == nn.InstanceNorm2d
        if input_nc is None:
            input_nc = outer_nc
        downconv = nn.Conv2d(input_nc, inner_nc, kernel_size=4,
                             stride=2, padding=1, bias=use_bias)
        downrelu = nn.LeakyReLU(0.2, True)
        downnorm = norm_layer(inner_nc)
        uprelu = nn.ReLU(True)
        upnorm = norm_layer(outer_nc)

        if outermost:
            upconv = nn.ConvTranspose2d(inner_nc * 2, outer_nc,
                                        kernel_size=4, stride=2,
                                        padding=1)
            down = [downconv]
            up = [uprelu, upconv, nn.Tanh()]
            model = down + [submodule] + up
        elif innermost:
            upconv = nn.ConvTranspose2d(inner_nc, outer_nc,
                                        kernel_size=4, stride=2,
                                        padding=1, bias=use_bias)
            down = [downrelu, downconv]
            up = [uprelu, upconv, upnorm]
            model = down + up
        else:
            upconv = nn.ConvTranspose2d(inner_nc * 2, outer_nc,
                                        kernel_size=4, stride=2,
                                        padding=1, bias=use_bias)
            down = [downrelu, downconv, downnorm]
            up = [uprelu, upconv, upnorm]

            if use_dropout:
                model = down + [submodule] + up + [nn.Dropout(0.5)]
            else:
                model = down + [submodule] + up

        self.model = nn.Sequential(*model)

    def forward(self, x):
        if self.outermost:
            return self.model(x)
        else:   # add skip connections
            return torch.cat([x, self.model(x)], 1)


class LineartAnimeDetector:
    model_dir = os.path.join(models_path, "lineart_anime")

    def __init__(self):
        self.model = None
        self.device = devices.get_device_for("controlnet")

    def load_model(self):
        remote_model_path = "https://huggingface.co/lllyasviel/Annotators/resolve/main/netG.pth"
        modelpath = os.path.join(self.model_dir, "netG.pth")
        if not os.path.exists(modelpath):
            from basicsr.utils.download_util import load_file_from_url
            load_file_from_url(remote_model_path, model_dir=self.model_dir)
        norm_layer = functools.partial(nn.InstanceNorm2d, affine=False, track_running_stats=False)
        net = UnetGenerator(3, 1, 8, 64, norm_layer=norm_layer, use_dropout=False)
        ckpt = torch.load(modelpath)
        for key in list(ckpt.keys()):
            if 'module.' in key:
                ckpt[key.replace('module.', '')] = ckpt[key]
                del ckpt[key]
        net.load_state_dict(ckpt)
        net.eval()
        self.model = net.to(self.device)

    def unload_model(self):
        if self.model is not None:
            self.model.cpu()

    def __call__(self, input_image):
        if self.model is None:
            self.load_model()
        self.model.to(self.device)

        H, W, C = input_image.shape
        Hn = 256 * int(np.ceil(float(H) / 256.0))
        Wn = 256 * int(np.ceil(float(W) / 256.0))
        img = cv2.resize(input_image, (Wn, Hn), interpolation=cv2.INTER_CUBIC)
        with torch.no_grad():
            image_feed = torch.from_numpy(img).float().to(self.device)
            image_feed = image_feed / 127.5 - 1.0
            image_feed = rearrange(image_feed, 'h w c -> 1 c h w')

            line = self.model(image_feed)[0, 0] * 127.5 + 127.5
            line = line.cpu().numpy()

            line = cv2.resize(line, (W, H), interpolation=cv2.INTER_CUBIC)
            line = line.clip(0, 255).astype(np.uint8)
            return line