Auto annotate images using TensorFlow object detection

1. The entire URL of the file you are using

https://github.com/tensorflow/models/tree/master/research/object_detection

2. Describe the feature you request

Hello this is my first contributing. I've implemented a Python class to get the inference of a frozen model and use this to create a xml file that label the images. With this, we can use a simple pre-treined model in our dataset to create new labels, accelerating the process of labelling images.
I was thinking that as great if we could work together in this code to implement a official version for the community! It's possible?

3. Additional context

I've wrote an article on medium about this: https://medium.com/analytics-vidhya/auto-annotate-images-for-tensorflow-object-detection-19b59f31c4d9

4. Are you willing to contribute it? definelly!