This is my EE551 python individual project | Author: HAO CAO
As a beginner python learner, it would be interesting to combine the program with my hobbie: painting.Van gogh is my idol, his painting named The Starry Night is my favourite, we cannot paint a mastwepiece like this, but maybe we can creat other pieces with the help of Python! It's amazing!
- Configure the running environment using the VGG16 model
- Build style transfer platform with keras
- deep learning algorithm
- Let the computer learn the input image style
- Style is digitized by deep learning algorithm
- The target image was processed iteratively to obtain similar style images
- Installate some dependent libraries with Anaconda
- Configure the operating environment
- Code designing about dimensions of the generated picture and util function to convert a tensor into a valid image.
- Run the program on Anaconda and input the run statement, get what we want.
- HAO CAO
pip install keras pip install h5py pip install numpy pip install scipy pip install tensorflow pip install matplotlib pip install Pillow
download the VGG16 model from Internet
please refer to neural_style_transfer.py.
run the program on the Anaconda and input: python neural_style_transfer.py ./stevens.jpg ./starry_night.jpg ./stevens_t
Introduce pictures and put them and neural_style_transfer.py.in the same folder.
This is Van gogh's The Starry Night:
This is our stevens School:
After 10 times iteration: we got them: