172 followers
..by groups at Uni of HK and UPenn. We show by a blind "turing test" style comparison that manual and automated labellings are hard to discriminate between, which is highly encouraging! Code is available on github and paper is available here - https://t