Hi

I'm considering a PhD in computer vision but have some concerns and I wondered if anyone would have any input.

Firstly, I've studied some introductory computer vision from my MSc in AI, but only a little. I did my thesis on something else. My undergraduate degree is mathematics (mostly applied). Does anyone know if a PhD would be the best way forwards to a computer vision career? I know many computer vision job adverts I can find are requiring a fair whack of experience or knowledge; some requiring a PhD but not all. Is it unlikely I will find a place in industry to pick up the experience? It feels like a PhD would be the best way to pick up the experience at this point but I'm not entirely sure...it's the old problem of picking up the experience in the first place I suppose.

Secondly, I'm interested in computer vision applied to autonomous vehicles (AV) but it isn't a must for me. Am I being too inflexible only considering PhDs which combine computer vision and AVs? If, for example, I took a PhD combining computer vision and deep learning in the marine/environmental domain, would it be reasonable to transfer to other industries (such as AVs) later on in life? Or should one wait around until finding the best possible PhD project?

I'd really appreciate any thoughts, especially if someone reading this already has experience in these topics. Thank you!