Boosting low-resolution images via the application of a deep network alongside an adversarial network (Generative Adversarial Networks) to generate high-resolution counterparts.

* Implemented a network architecture consisting of 16 Residual blocks to enhance the learning capabilities of the GAN.
* Utilized PixelShuffler x2, integrating two sub-pixel CNNs within the Generator for efficient feature map upscaling.
* Adopted the Parameterized Rectified Linear Unit (PReLU) activation function, enhancing the model's ability to learn adaptive coefficients for negative inputs.
* Employed convolutional layers with specifications like k3n64s1, indicating a kernel size of 3, 64 channels, and strides of 1, optimizing feature extraction.
* Incorporated a Perceptual loss function comprising Content (Reconstruction) loss and Adversarial loss to guide the training process effectively.

* We process the HR(High Resolution) images to get down-sampled LR(Low Resolution) images. Now we have both HR 
  and LR images for training data set.
* We pass LR images through Generator which up-samples and gives SR(Super Resolution) images.
* We use a discriminator to distinguish the HR images and back-propagate the GAN loss to train the discriminator
  and the generator.
* As a result of this, the generator learns to produce more and more realistic images(High Resolution images) as 
  it trains.

You will need the following to run the above:
Python 3.5.8 
tensorflow 1.16.0
keras 2.2.4
numpy 1.10.4
matplotlib, skimage, opencv-python

* Used COCO data set 2017. It is around 18GB having images of different dimensions.
* After above step you have High Resolution images. Now you have to get Low Resolution images which you can get by down 
  scaling High Res  images. I used down scale = 4. So Low resolution image of size 96 we will get.

 * Training:
   already done and weights are provided
    
 * Testing:
        1. Test Model- Here you can test the model by providing High Res images. It will process to get resulting LR images and then will generate SR images.
           And then will save output file comprising of all Low Res, Super Res and High Res images.
           Run following command to test model:
           > python test.py -ihr='./data_hr/' -o='./output/' -m='./model/gen_model3000.h5' -n-1 -t='test_model'
           
           
        2. Test Low Res images- This option directly take Low Res images and give resulting High Res images.
           > python test.py -ilr='./data_lr/' -o='./output/' -m='./model/gen_model3000.h5' -n=1 -t='test_lr_images'
           For more help run:

        3. For more info
           python test.py -h

PS:- ./data/ folder can be made in the root directory for use

*  GAN training can be challenging, especially with deep networks, but incorporating residual blocks alleviates some of these difficulties.
*  Understanding Perceptual loss simplifies training and opens doors to applications like Image Style Transfer and Photo Realistic Style Transfer.
*  Acquiring suitable data posed a challenge, emphasizing the importance of careful selection and preprocessing.
*  It's advisable to use images with uniform width and height for smoother processing.