Understanding how variations in data impact results

[wentdot]

I am curious about how to connect specific features of my data to my results.

The "rules of thumb" make sense - including a variety of different poses, expressions, lighting, etc. But supposing the data set is necessarily flawed in some way - let's say half the data set is from before the person got plastic surgery on their nose, or from 5-10 years earlier. How does the disparity in the data manifest in the swap?

Are features that are inconsistent in the data necessarily averaged (and thus blurred?) in the result?

Suppose half of the pictures feature a mole that was removed - should we expect a fainter mole in the swap?

[abigflea]

This is a curious question. I like it.

For the mole question, I would suspect it would do like beards and glasses, depending on frame and lighting, it would appear and disappear, maybe hang around faded.

For the entirety of your question.. who knows. Could be an interesting test.