What is the correct way to train the face detection model?

I have tens of thousands of photos which include the hundreds of people many hundreds of times each. Many of their faces in these photos are quite poor quality (obscured, dark, far away). When I start recognition it is often quite poor but once I have a few high quality faces in the model it seems to work fairly well. However as I accept more of the poor quality faces it find then the recognition gets worse (I assume because dark, far way, poor quality faces look very similar).

What is the intended workflow? It seems like I should accept all recognised faces and use them in the training. Instead should I only add high quality faces to the face tags and tag poor quality ones differently? It would be good to be able to mark which faces you want used in the training.

Thanks for any help.