NS之VGG(Keras):基于Keras的VGG16实现之《复仇者联盟3》灭霸图像风格迁移设计(A Neural Algorithm of Artistic Style)
导读
通过代码设计,基于Keras的VGG16实现A Neural Algorithm of Artistic Style之灭霸风格迁移设计
目录
灭霸风格素材
1、对我的偶像胡歌,下手……
- 22044672/58889256 [==========>...................] - ETA: 1:53
- 22077440/58889256 [==========>...................] - ETA: 1:53
- 22093824/58889256 [==========>...................] - ETA: 1:53
- 22126592/58889256 [==========>...................] - ETA: 1:52
- 22142976/58889256 [==========>...................] - ETA: 1:52
- 22183936/58889256 [==========>...................] - ETA: 1:52
- ……
- 58826752/58889256 [============================>.] - ETA: 0s
- 58859520/58889256 [============================>.] - ETA: 0s
- 58875904/58889256 [============================>.] - ETA: 0s
- 58892288/58889256 [==============================] - 250s 4us/step
- Model loaded.
-
- Start of iteration 0
- Current loss value: 2254528800.0
- Image saved as F:\File_Python\Python_example\fast_style_transfer_master\examples\results_at_iteration_0.png
- Iteration 0 completed in 252s
- Start of iteration 1
- Current loss value: 1771042400.0
- Image saved as F:\File_Python\Python_example\fast_style_transfer_master\examples\results_at_iteration_1.png
- Iteration 1 completed in 238s
- Start of iteration 2
2、其他风格迁移设计
- Keras:基于Keras的VGG16实现A Neural Algorithm of Artistic Style
- from keras.preprocessing.image import load_img, img_to_array
- import numpy as np
- from scipy.misc import imsave
- from scipy.optimize import fmin_l_bfgs_b
- import time
- from keras.applications.vgg16 import VGG16
- from keras.applications.vgg16 import preprocess_input
- from keras import backend as K
- import os
-
- References
- - [A Neural Algorithm of Artistic Style](http://arxiv.org/abs/1508.06576)
- '''
- base_image_path=r'F:\File_Python\Python_example\fast_style_transfer_master\examples\content\20180214(21).jpg'
- style_reference_image_path=r'F:\File_Python\Python_example\fast_style_transfer_master\examples\style\mieba02.jpg'
- result_prefix=r'F:\File_Python\Python_example\fast_style_transfer_master\examples\results'
- iterations=10
- content_weight=0.025
- style_weight=1.0
- total_variation_weight=1.0
网站声明:如果转载,请联系本站管理员。否则一切后果自行承担。
加入交流群
请使用微信扫一扫!