In this assignment you have to build a model for super-resolution using tf.keras. You have to train a CNN to upscale small image patches, then use the trained model to upscale full images. To implement such kind of model, you should take a look at the following classes and methods: Sequential model, Functional API, Conv2D, cv2.cvtColor (BGR2YCrCb, YCrCb2RGB, YCrCb2BGR).
You can use only these 3rd party packages: cv2, pandas, keras, matplotlib, numpy, sklearn, skimage, tensorflow.