|
import numpy as np |
|
import torch |
|
import torch.nn as nn |
|
import gradio as gr |
|
from PIL import Image |
|
import torchvision.transforms as transforms |
|
|
|
norm_layer = nn.InstanceNorm2d |
|
|
|
class ResidualBlock(nn.Module): |
|
def __init__(self, in_features): |
|
super(ResidualBlock, self).__init__() |
|
|
|
conv_block = [ nn.ReflectionPad2d(1), |
|
nn.Conv2d(in_features, in_features, 3), |
|
norm_layer(in_features), |
|
nn.ReLU(inplace=True), |
|
nn.ReflectionPad2d(1), |
|
nn.Conv2d(in_features, in_features, 3), |
|
norm_layer(in_features) |
|
] |
|
|
|
self.conv_block = nn.Sequential(*conv_block) |
|
|
|
def forward(self, x): |
|
return x + self.conv_block(x) |
|
|
|
|
|
class Generator(nn.Module): |
|
def __init__(self, input_nc, output_nc, n_residual_blocks=9, sigmoid=True): |
|
super(Generator, self).__init__() |
|
|
|
|
|
model0 = [ nn.ReflectionPad2d(3), |
|
nn.Conv2d(input_nc, 64, 7), |
|
norm_layer(64), |
|
nn.ReLU(inplace=True) ] |
|
self.model0 = nn.Sequential(*model0) |
|
|
|
|
|
model1 = [] |
|
in_features = 64 |
|
out_features = in_features*2 |
|
for _ in range(2): |
|
model1 += [ nn.Conv2d(in_features, out_features, 3, stride=2, padding=1), |
|
norm_layer(out_features), |
|
nn.ReLU(inplace=True) ] |
|
in_features = out_features |
|
out_features = in_features*2 |
|
self.model1 = nn.Sequential(*model1) |
|
|
|
model2 = [] |
|
|
|
for _ in range(n_residual_blocks): |
|
model2 += [ResidualBlock(in_features)] |
|
self.model2 = nn.Sequential(*model2) |
|
|
|
|
|
model3 = [] |
|
out_features = in_features//2 |
|
for _ in range(2): |
|
model3 += [ nn.ConvTranspose2d(in_features, out_features, 3, stride=2, padding=1, output_padding=1), |
|
norm_layer(out_features), |
|
nn.ReLU(inplace=True) ] |
|
in_features = out_features |
|
out_features = in_features//2 |
|
self.model3 = nn.Sequential(*model3) |
|
|
|
|
|
model4 = [ nn.ReflectionPad2d(3), |
|
nn.Conv2d(64, output_nc, 7)] |
|
if sigmoid: |
|
model4 += [nn.Sigmoid()] |
|
|
|
self.model4 = nn.Sequential(*model4) |
|
|
|
def forward(self, x, cond=None): |
|
out = self.model0(x) |
|
out = self.model1(out) |
|
out = self.model2(out) |
|
out = self.model3(out) |
|
out = self.model4(out) |
|
|
|
return out |
|
|
|
model = Generator(3, 1, 3) |
|
model.load_state_dict(torch.load('model.pth', map_location=torch.device('cpu'))) |
|
model.eval() |
|
|
|
transforms_r = [transforms.Resize(int(opt.size), Image.BICUBIC), |
|
transforms.ToTensor()] |
|
transforms.Compose([transforms.Resize(int(opt.size), Image.BICUBIC), transforms.ToTensor()]) |
|
|
|
def predict(input_img): |
|
input_img = Image.open(input_img) |
|
ratio = 256.0 / input_img.size[1] |
|
input_img = input_img.resize((int(ratio*input_img.size[0]),256), Image.ANTIALIAS) |
|
|
|
|
|
return input_img |
|
|
|
title="informative-drawings" |
|
description="Gradio Demo for line drawing generation. " |
|
article = "<p style='text-align: center'><a href='https://film-net.github.io/' target='_blank'>FILM: Frame Interpolation for Large Motion</a> | <a href='https://github.com/google-research/frame-interpolation' target='_blank'>Github Repo</a></p>" |
|
examples=[['cat.png']] |
|
|
|
|
|
iface = gr.Interface(predict, gr.inputs.Image(type='filepath'), "image", title=title,description=description,article=article,examples=examples) |
|
|
|
iface.launch() |
