This is a surprisingly easy mistake to make.<p>Background: I was one of the original engineers on the Goal Line Technology system now used by the vast majority of professional leagues.<p>Heads are surprisingly ball shaped from a a lot of camera angles, and the male balding pattern conspires to give a good impression of different football graphical designs.<p>When we were developing the system, we thought we had ball detection pretty nailed, and we never had false positives from any of the heads of engineers on the team. Then, our boss came to test the system ... As far as our algorithms were concerned, his head was exactly a football, just as in the article! Highly embarrassing, but everyone saw the funny side. It gave us better data and inspired a few more robustness checks that ended up being crucial later on! :)
And it's totally okay! Because nobody died, we all get a nice laugh, and the AI will be tweaked to make it better. It's the kind of move-fast-and-break-things that I can get behind.
I am not seeing an obvious scenario that makes it switch. I thought maybe it was when the ball "joined" with a player but it does it when its in motion too. That is just completely broken.<p>We often hear that the AI is more accurate than humans in some image matching, but it also clearly makes mistakes that no human would make. Well here is a great example of such a mistake which a human wouldn't be making.
It's crazy thinking about the amount of technology that went into the infamous "Fox Puck" for tracking and display (where they literally cut a puck in half to embed tech and coordinate with an array of receivers in the arena) to what can now be done by the cameras themselves.<p><a href="https://en.m.wikipedia.org/wiki/FoxTrax" rel="nofollow">https://en.m.wikipedia.org/wiki/FoxTrax</a><p>I suspect in this instance it wasn't just the bald head, but the particular near-sunset lighting that was a contributing factor.
It's kind of mystifying why a company would go to the trouble of having AI-based soccer camera to begin with.<p>I can't imagine that the training and maintenance of such a system could be cheaper than the pay of a single camera person and I don't remember ordinary camera people having much trouble following the ball in soccer.<p>Which is to say, while current ML/AI may work well for some things, for other things, it will be abandoned as a fad unless there's some fundamental improvement.
YouTube video for those interested, music overlay warning though.<p><a href="https://youtu.be/9zoJP2FkpgU" rel="nofollow">https://youtu.be/9zoJP2FkpgU</a>
Reminds me of an incident I read about in Hackers by Steven Levy, where a ball-catching robot in the MIT AI Lab (in the 60s or 70s) mistook Prof. Marvin Minsky's bald head for a ping-pong ball.
All sorts of edge cases here to handle. Its not just a bald head.<p>What happens if there are multiple balls? What happens if the ball is redesigned with a different pattern? What happens if someone wears a shirt with a picture of a ball?<p>Does it make sense to center the camera on the ball? Does it make sense to also show where the ball can go?<p>I wonder how it's handled by people currently? I suppose it can be handled with more training data/labelling instead of custom coding.<p>We are looking are replacing the camera person job with AI just as we replaced other people's jobs with programming earlier and just as we replaced other people's jobs with AI more recently.
I wonder what went wrong here. Perhaps it wasn't trained for these lighting conditions? I can't tell if they use multiple cameras around the stadium or not.
Wow, so this companies tech also does automated - "highlights, replays, and adverts"<p><a href="https://www.pixellot.tv/" rel="nofollow">https://www.pixellot.tv/</a>
Why are so cheap to replace a cameraman by an AI?
They could leave the camera control from remote to someone in a low income country, would make a better job for peanuts.
While it's hilarious, I wonder why they didn't keep a human backup available just in case. Anyway, I think the solution is clear: the ref must be replaced by AI.