diff options
| -rw-r--r-- | __pycache__/hyperparameters.cpython-38.pyc | bin | 341 -> 332 bytes | |||
| -rw-r--r-- | __pycache__/losses.cpython-38.pyc | bin | 4409 -> 4546 bytes | |||
| -rw-r--r-- | hyperparameters.py | 6 | ||||
| -rw-r--r-- | losses.py | 37 | ||||
| -rw-r--r-- | main.py | 12 | ||||
| -rw-r--r-- | save.jpg | bin | 39903 -> 46064 bytes |
6 files changed, 40 insertions, 15 deletions
diff --git a/__pycache__/hyperparameters.cpython-38.pyc b/__pycache__/hyperparameters.cpython-38.pyc Binary files differindex e1a90bd4..7d32eefa 100644 --- a/__pycache__/hyperparameters.cpython-38.pyc +++ b/__pycache__/hyperparameters.cpython-38.pyc diff --git a/__pycache__/losses.cpython-38.pyc b/__pycache__/losses.cpython-38.pyc Binary files differindex 71b86245..66e565ec 100644 --- a/__pycache__/losses.cpython-38.pyc +++ b/__pycache__/losses.cpython-38.pyc diff --git a/hyperparameters.py b/hyperparameters.py index b03db017..460543dc 100644 --- a/hyperparameters.py +++ b/hyperparameters.py @@ -9,17 +9,17 @@ Number of epochs. If you experiment with more complex networks you might need to increase this. Likewise if you add regularization that slows training. """ -num_epochs = 5000 +num_epochs = 1000 """ A critical parameter that can dramatically affect whether training succeeds or fails. The value for this depends significantly on which optimizer is used. Refer to the default learning rate parameter """ -learning_rate = 3e-2 +learning_rate = 1e-2 momentum = 0.01 alpha = 1e-2 -beta = 1e-5 +beta = 1e-4 @@ -1,8 +1,10 @@ import tensorflow as tf +import numpy as np from tensorflow.keras.layers import \ Conv2D, AveragePooling2D from skimage import transform import hyperparameters as hp + class YourModel(tf.keras.Model): """ Your own neural network model. """ @@ -11,17 +13,26 @@ class YourModel(tf.keras.Model): self.content_image = transform.resize(content_image, tf.shape(style_image), anti_aliasing=True) self.content_image = tf.expand_dims(self.content_image, axis=0) + print(self.content_image) #perhaps consider cropping to avoid distortion self.style_image = transform.resize(style_image, tf.shape(style_image), anti_aliasing=True) self.style_image = tf.expand_dims(self.style_image, axis=0) - self.x = tf.Variable(tf.expand_dims(tf.random.uniform(tf.shape(content_image)), axis=0), trainable=True) + #self.x = tf.Variable(initial_value = self.content_image.numpy().astype(np.float32), trainable=True) + self.x = tf.Variable(initial_value = np.random.rand(self.content_image.shape[0], + self.content_image.shape[1], self.content_image.shape[2], self.content_image.shape[3]).astype(np.float32), trainable=True) + self.alpha = hp.alpha self.beta = hp.beta - print(self.x.shape) + self.photo_layers = None + self.art_layers = None + - print(self.content_image.shape, self.style_image.shape) + + #(self.x.shape) + + #print(self.content_image.shape, self.style_image.shape) self.optimizer = tf.keras.optimizers.Adam() @@ -88,14 +99,20 @@ class YourModel(tf.keras.Model): return x, layers def loss_fn(self, p, a, x): - _, photo_layers = self.call(p) - _, art_layers = self.call(a) - _, input_layers = self.call(x) - - content_l = self.content_loss(photo_layers, input_layers) - style_l = self.style_loss(art_layers, input_layers) + # print(p) + if(self.photo_layers == None): + _, self.photo_layers = self.call(p) + # print(a) + if(self.art_layers == None): + _, self.art_layers = self.call(a) + # print(x) + _, input_layers = self.call(x) + + + content_l = self.content_loss(self.photo_layers, input_layers) + style_l = self.style_loss(self.art_layers, input_layers) # Equation 7 - return (self.alpha * content_l) + (self.beta * style_l) + return (self.alpha * content_l) + (self.beta * style_l) def content_loss(self, photo_layers, input_layers): L_content = tf.constant(0.0) @@ -1,7 +1,9 @@ import os import sys import argparse +import cv2 import tensorflow as tf +from skimage import transform import hyperparameters as hp from losses import YourModel @@ -50,9 +52,15 @@ def main(): print('this is',ARGS.content) content_image = imread(ARGS.content) - content_image = np.resize(content_image, (255, 255, 3)) + + style_image = imread(ARGS.style) - style_image = np.resize(style_image, (255, 255, 3)) + cv2.imshow('hi1', style_image) + cv2.waitKey(0) + + style_image = transform.resize(style_image, content_image.shape) + cv2.imshow('hi2', style_image) + cv2.waitKey(0) my_model = YourModel(content_image=content_image, style_image=style_image) my_model.vgg16.build([1, 255, 255, 3]) my_model.vgg16.load_weights('vgg16_imagenet.h5', by_name=True) Binary files differ |
