Object Deteccion in surgical videos -- 2

The project is about recognition of some tools in a surgical video. I have a database with videos (train and test datasets), using the train dataset and faster R-CNN a person should create architecture that: draw bounding boxes in the object and classify this object.
The restriction for this project is use Faster R-CNN and use Python as language. The presentation of solution dont need to be special, just the code in a notebook is ok.