/Neural_Style_Transfer

Using Pytorch, keras, Tensorflow and FastAI for implementing neural style transfer on given image dataset.

Primary LanguageJupyter NotebookApache License 2.0Apache-2.0

Neural_Style_Transfer

Using Pytorch, keras, Tensorflow and FastAI for implementing neural style transfer on given image dataset.

PYTORCH

Using pretrained VGG19 model on pytorch neural style transfer was possible

  1. We first initialized the runtime on colab to GPU for faster processing
  2. Load the torch libraries and pretrained VGG19 model
  3. Get a content and a style image after defining the function to load the Image using PIL.Image
  4. Convert images to tensors
  5. Gram matrix method is done on the images
  6. training and validation steps are computed.

References:

https://github.com/smara97/Machine-Learning-Projects/blob/master/Neural%20Style%20Transfer/neural-style-transfer.ipynb

##TODO 1)Keras 2) Tensorflow 3) fastAI

feel free to use and modify my repo and do let me know if any changes could be made. Cheers!!