I'm trying to help you understand what "ground truth" means.

If, as it seems in the article, they are using COCO to establish ground truth, i.e. what COCO says is correct, then whatever COCO comes up with is, by definition "correct". It is, in effect, the answer, the measuring stick, the scoring card. Now what you're hinting at is that, in this instance, that's a really bad way to establish ground truth. I agree. But that doesn't change what is and how we use ground truth.

Think of it another way:

- Your job is to pass a test.

- To pass a test you must answer a question correctly.

- The answer to that question has already been written down somewhere.

To pass the test does your answer need to be true, or does it need to match what is already written down?

When we do model evaluation the answer needs to match what is already written down.

So, it sounds like you’re saying that the ML field has hijacked the well-defined and -understood term “ground truth”, to mean something that should be similar, but which is fundamentally unrelated, and in cases like this is in no way similar. Even what it is to be “correct” is damaged.

I am willing to accept that this is how they are using the terms; but it distresses me. They should choose appropriate terms rather than misappropriating existing terms.

(Your address example I still don’t get, because I expect your model to do some massaging to match custom, so I wouldn’t consider an Address Line 1 of “J Smith 123 Example St” with empty Address Line 2 to be true or correct.)

You're trying so hard not to learn something new in this thread, that it's almost impressive.