TensorFlow Lite를 사용한 예술적 스타일 전이

import tensorflow as tf
print(tf.__version__)

2022-12-14 20:34:17.209224: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory
2022-12-14 20:34:17.209325: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory
2022-12-14 20:34:17.209335: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
2.11.0

import IPython.display as display

import matplotlib.pyplot as plt
import matplotlib as mpl
mpl.rcParams['figure.figsize'] = (12,12)
mpl.rcParams['axes.grid'] = False

import numpy as np
import time
import functools

content_path = tf.keras.utils.get_file('belfry.jpg','https://storage.googleapis.com/khanhlvg-public.appspot.com/arbitrary-style-transfer/belfry-2611573_1280.jpg')
style_path = tf.keras.utils.get_file('style23.jpg','https://storage.googleapis.com/khanhlvg-public.appspot.com/arbitrary-style-transfer/style23.jpg')

style_predict_path = tf.keras.utils.get_file('style_predict.tflite', 'https://tfhub.dev/google/lite-model/magenta/arbitrary-image-stylization-v1-256/int8/prediction/1?lite-format=tflite')
style_transform_path = tf.keras.utils.get_file('style_transform.tflite', 'https://tfhub.dev/google/lite-model/magenta/arbitrary-image-stylization-v1-256/int8/transfer/1?lite-format=tflite')

Downloading data from https://storage.googleapis.com/khanhlvg-public.appspot.com/arbitrary-style-transfer/belfry-2611573_1280.jpg
458481/458481 [==============================] - 0s 0us/step
Downloading data from https://storage.googleapis.com/khanhlvg-public.appspot.com/arbitrary-style-transfer/style23.jpg
108525/108525 [==============================] - 0s 0us/step
Downloading data from https://tfhub.dev/google/lite-model/magenta/arbitrary-image-stylization-v1-256/int8/prediction/1?lite-format=tflite
2828838/2828838 [==============================] - 0s 0us/step
Downloading data from https://tfhub.dev/google/lite-model/magenta/arbitrary-image-stylization-v1-256/int8/transfer/1?lite-format=tflite
284398/284398 [==============================] - 0s 0us/step

# Function to load an image from a file, and add a batch dimension.
def load_img(path_to_img):
  img = tf.io.read_file(path_to_img)
  img = tf.io.decode_image(img, channels=3)
  img = tf.image.convert_image_dtype(img, tf.float32)
  img = img[tf.newaxis, :]

  return img

# Function to pre-process by resizing an central cropping it.
def preprocess_image(image, target_dim):
  # Resize the image so that the shorter dimension becomes 256px.
  shape = tf.cast(tf.shape(image)[1:-1], tf.float32)
  short_dim = min(shape)
  scale = target_dim / short_dim
  new_shape = tf.cast(shape * scale, tf.int32)
  image = tf.image.resize(image, new_shape)

  # Central crop the image.
  image = tf.image.resize_with_crop_or_pad(image, target_dim, target_dim)

  return image

# Load the input images.
content_image = load_img(content_path)
style_image = load_img(style_path)

# Preprocess the input images.
preprocessed_content_image = preprocess_image(content_image, 384)
preprocessed_style_image = preprocess_image(style_image, 256)

print('Style Image Shape:', preprocessed_style_image.shape)
print('Content Image Shape:', preprocessed_content_image.shape)

Style Image Shape: (1, 256, 256, 3)
Content Image Shape: (1, 384, 384, 3)

def imshow(image, title=None):
  if len(image.shape) > 3:
    image = tf.squeeze(image, axis=0)

  plt.imshow(image)
  if title:
    plt.title(title)

plt.subplot(1, 2, 1)
imshow(preprocessed_content_image, 'Content Image')

plt.subplot(1, 2, 2)
imshow(preprocessed_style_image, 'Style Image')

# Function to run style prediction on preprocessed style image.
def run_style_predict(preprocessed_style_image):
  # Load the model.
  interpreter = tf.lite.Interpreter(model_path=style_predict_path)

  # Set model input.
  interpreter.allocate_tensors()
  input_details = interpreter.get_input_details()
  interpreter.set_tensor(input_details[0]["index"], preprocessed_style_image)

  # Calculate style bottleneck.
  interpreter.invoke()
  style_bottleneck = interpreter.tensor(
      interpreter.get_output_details()[0]["index"]
      )()

  return style_bottleneck

# Calculate style bottleneck for the preprocessed style image.
style_bottleneck = run_style_predict(preprocessed_style_image)
print('Style Bottleneck Shape:', style_bottleneck.shape)

Style Bottleneck Shape: (1, 1, 1, 100)
INFO: Created TensorFlow Lite XNNPACK delegate for CPU.

# Run style transform on preprocessed style image
def run_style_transform(style_bottleneck, preprocessed_content_image):
  # Load the model.
  interpreter = tf.lite.Interpreter(model_path=style_transform_path)

  # Set model input.
  input_details = interpreter.get_input_details()
  interpreter.allocate_tensors()

  # Set model inputs.
  interpreter.set_tensor(input_details[0]["index"], preprocessed_content_image)
  interpreter.set_tensor(input_details[1]["index"], style_bottleneck)
  interpreter.invoke()

  # Transform content image.
  stylized_image = interpreter.tensor(
      interpreter.get_output_details()[0]["index"]
      )()

  return stylized_image

# Stylize the content image using the style bottleneck.
stylized_image = run_style_transform(style_bottleneck, preprocessed_content_image)

# Visualize the output.
imshow(stylized_image, 'Stylized Image')

# Calculate style bottleneck of the content image.
style_bottleneck_content = run_style_predict(
    preprocess_image(content_image, 256)
    )

# Define content blending ratio between [0..1].
# 0.0: 0% style extracts from content image.
# 1.0: 100% style extracted from content image.
content_blending_ratio = 0.5

# Blend the style bottleneck of style image and content image
style_bottleneck_blended = content_blending_ratio * style_bottleneck_content \
                           + (1 - content_blending_ratio) * style_bottleneck

# Stylize the content image using the style bottleneck.
stylized_image_blended = run_style_transform(style_bottleneck_blended,
                                             preprocessed_content_image)

# Visualize the output.
imshow(stylized_image_blended, 'Blended Stylized Image')