def stylize(args):
device = torch.device("cuda" if args.cuda else "cpu")
content_image = utils.load_image(args.content_image,
scale=args.content_scale)
content_transform = transforms.Compose(
[transforms.ToTensor(),
transforms.Lambda(lambda x: x.mul(255))])
content_image = content_transform(content_image)
content_image = content_image.unsqueeze(0).to(device)
if args.model.endswith(".onnx"):
output = stylize_onnx_caffe2(content_image, args)
else:
with torch.no_grad():
style_model = TransformerNet()
state_dict = torch.load(args.model)
# remove saved deprecated running_* keys in InstanceNorm from the checkpoint
for k in list(state_dict.keys()):
if re.search(r'in\d+\.running_(mean|var)$', k):
del state_dict[k]
style_model.load_state_dict(state_dict)
style_model.to(device)
if args.export_onnx:
assert args.export_onnx.endswith(
".onnx"), "Export model file should end with .onnx"
output = torch.onnx._export(style_model, content_image,
args.export_onnx).cpu()
else:
output = style_model(content_image).cpu()
utils.save_image(args.output_image, output[0])
def tale():
data = json.loads(request.get_data(), encoding='utf-8')
story_pk = data['story_pk']
paths = data['imagePaths']
cnt = 0
for path in paths:
# image load from url
image_url = IMAGE_SERVER_URL + '/media/' +path
res = requests.get(image_url)
img = Image.open(BytesIO(res.content))
img = np.asarray(img)
for i in range(6):
with tf.Graph().as_default():
transformer = style_transfer_tester.StyleTransferTester(
img, 'neural_style/fast_neural_style/' + STYLE_PATH[i]
)
output = transformer.test()
result_path = str(story_pk) + '_' + str(cnt) + '_'+ str(i) + '.jpg'
utils.save_image(output, 'static/' + result_path)
cnt = cnt + 1
params = {
'story_pk' : story_pk,
'tale': ['아직', '준비중인', '컨텐츠', '입니다', 'ㅎㅎ']
}
requests.post('http://127.0.0.1:8000/storys/editable/', data = params)
return 'ok'
def stylize(input_filepath, output_filepath, checkpoint, content_scale=None):
content_image = utils.load_image(input_filepath, scale=content_scale)
content_transform = transforms.Compose(
[transforms.ToTensor(),
transforms.Lambda(lambda x: x.mul(255))])
content_image = content_transform(content_image)
content_image = content_image.unsqueeze(0)
if CUDA:
content_image = content_image.cuda()
content_image = Variable(content_image, volatile=True)
style_model = TransformerNet()
style_model.load_state_dict(torch.load(checkpoint))
if CUDA:
style_model.cuda()
output = style_model(content_image)
if CUDA:
output = output.cpu()
output_data = output.data[0]
utils.save_image(output_filepath, output_data)
def style():
# Get image url from json
path = json.loads(request.get_data(), encoding='utf-8')
image_url = path['url']
# Image load from url
res = requests.get(image_url)
img = Image.open(BytesIO(res.content))
img = np.asarray(img)
img_extension = image_url[-4:]
img_extension_path = tf.keras.utils.get_file('image'+img_extension,
origin=image_url)
for i in range(1,6):
with tf.Graph().as_default():
# run neural network
transformer = style_transfer_tester.StyleTransferTester(
img, 'neural_style/fast_neural_style/wave.ckpt'
)
output = transformer.test()
# save result
result_path = 'asdf3.jpg'
utils.save_image(output, 'static/'+str(i)+'_1'+result_path)
with tf.Graph().as_default():
# run neural network
transformer = style_transfer_tester.StyleTransferTester(
img, 'neural_style/fast_neural_style/udnie.ckpt'
)
output = transformer.test()
# save result
result_path = 'asdf3.jpg'
utils.save_image(output, 'static/'+str(i)+'_2'+result_path)
with tf.Graph().as_default():
caption_model = Image_caption()
result_cap , plot = caption_model(img_extension_path)
return result_cap
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
## put your original images in this directory
input_dir = Path('./inputs')
with torch.no_grad():
style_model = TransformerNet()
state_dict = torch.load('nausicaa3.model')
for k in list(state_dict.keys()):
if re.search(r'in\d+\.running_(mean|var)$', k):
del state_dict[k]
style_model.load_state_dict(state_dict)
style_model.to(device)
for imagefile in input_dir.glob('*.jpg'):
content_image = utils.load_image(imagefile, scale=None)
content_transform = transforms.Compose([
transforms.ToTensor(),
transforms.Lambda(lambda x: x.mul(255))
])
content_image = content_transform(content_image)
content_image = content_image.unsqueeze(0).to(device)
output = style_model(content_image).cpu()
# styled images will be saved to this directory
utils.save_image(f'./outputs/{imagefile.name}', output[0])