Recognizing objects in the photo (python)

You need to write and train a neural network that should recognize objects in the photo. At the input it should receive an image, at the output it should determine the class of the object. It is necessary for the neural network to determine the type of material:
- paper;
- cardboard;
- glass;
- plastic.
If there are several objects in the image, let it return "unknown".
It will be enough if it identifies materials with 20% accuracy, but it is necessary to be able to train it further by us.