eridgd / adain-tf Goto Github PK
View Code? Open in Web Editor NEWTensorFlow/Keras implementation of "Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization" https://arxiv.org/abs/1703.06868
License: MIT License
TensorFlow/Keras implementation of "Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization" https://arxiv.org/abs/1703.06868
License: MIT License
Hi, just wondering if it's possible to view the visual transfers as it progresses through training?
Not sure how this is done in tensorflow but in the original torch implementation you can view the training images . Any ideas?
Cheers
Hi Eridgd,
I saw most of the AdaIN-TF implementation used the normalised-vgg19 network as the encoder (include the original author)
Do you know the difference between normalised_vgg network and original VGG-19 network? And can we implement AdaIN-TF with the original vgg-19 network as encoder ?
Hi, do you have a decoder model you've already trained that I could test?
HI is there a CPU mode available? or is this GPU only?
Hey dude, just getting some errors when trying to train...any ideas?
(tf-keras) E:\SUGARBANK\ML\SOFTWARE\AdaIN-TF-master>python train.py --content-path E:\SUGARBANK\ML\SOFTWARE\train2014 --style-path E:\SUGARBANK\ML\ASSETS\Style --batch-size 8 --content-weight 1 --style-weight 1e-2 --tv-weight 0 --checkpoint E:\SUGARBANK\ML\AdaIN-TF-master\models\checkpoint --learning-rate 1e-4 --lr-decay 1e-5
Using TensorFlow backend.
2018-02-06 12:30:31.142079: I C:\tf_jenkins\workspace\rel-win\M\windows-gpu\PY\35\tensorflow\core\platform\cpu_feature_guard.cc:137] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX AVX2
2018-02-06 12:30:32.247307: I C:\tf_jenkins\workspace\rel-win\M\windows-gpu\PY\35\tensorflow\core\common_runtime\gpu\gpu_device.cc:1105] Found device 0 with properties:
name: GeForce GTX 980M major: 5 minor: 2 memoryClockRate(GHz): 1.1265
pciBusID: 0000:01:00.0
totalMemory: 8.00GiB freeMemory: 6.71GiB
2018-02-06 12:30:32.248644: I C:\tf_jenkins\workspace\rel-win\M\windows-gpu\PY\35\tensorflow\core\common_runtime\gpu\gpu_device.cc:1195] Creating TensorFlow device (/device:GPU:0) -> (device: 0, name: GeForce GTX 980M, pci bus id: 0000:01:00.0, compute capability: 5.2)
Traceback (most recent call last):
File "train.py", line 189, in
train()
File "train.py", line 128, in train
use_gram=args.gram)
File "E:\SUGARBANK\ML\SOFTWARE\AdaIN-TF-master\model.py", line 18, in init
self.build_model(vgg_weights)
File "E:\SUGARBANK\ML\SOFTWARE\AdaIN-TF-master\model.py", line 31, in build_model
self.vgg_model = vgg_from_t7(vgg_weights, target_layer='relu4_1')
File "E:\SUGARBANK\ML\SOFTWARE\AdaIN-TF-master\vgg_normalised.py", line 21, in vgg_from_t7
t7 = torchfile.load(t7_file, force_8bytes_long=False)
File "C:\Users\Sugar\AppData\Local\conda\conda\envs\tf-keras\lib\site-packages\torchfile.py", line 424, in load
return reader.read_obj()
File "C:\Users\Sugar\AppData\Local\conda\conda\envs\tf-keras\lib\site-packages\torchfile.py", line 370, in read_obj
obj._obj = self.read_obj()
File "C:\Users\Sugar\AppData\Local\conda\conda\envs\tf-keras\lib\site-packages\torchfile.py", line 385, in read_obj
k = self.read_obj()
File "C:\Users\Sugar\AppData\Local\conda\conda\envs\tf-keras\lib\site-packages\torchfile.py", line 386, in read_obj
v = self.read_obj()
File "C:\Users\Sugar\AppData\Local\conda\conda\envs\tf-keras\lib\site-packages\torchfile.py", line 370, in read_obj
obj._obj = self.read_obj()
File "C:\Users\Sugar\AppData\Local\conda\conda\envs\tf-keras\lib\site-packages\torchfile.py", line 387, in read_obj
obj[k] = v
TypeError: unhashable type: 'list'
Did you implement the Spatial Control? when I implement the Spatial control, I have some question。
Hi, I was wondering why you have calculated sum of squared
in style loss calculations:
m_loss = sse(d_mean, s_mean) / batch_size # normalized w.r.t. batch size
While in torch implementation MSE
is being used, this line:
self.mean_criterion = nn.MSECriterion() self.mean_loss = self.mean_criterion:forward(self.input_mean, self.target_mean)
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.