4 ms·
Fascinating. Also interesting to see the failure modes. Any human would quickly realize that the "boy doing backflip on wakeboard" is actually playing on a tram
by arjie 12y ago
Fascinating. Also interesting to see the failure modes. Any human would quickly realize that the "boy doing backflip on wakeboard" is actually playing on a trampoline. Or the "two young girls playing with legos toy". Great stuff!
- DavidSJ 12y agoI think I'm a human, but I did not realize that boy was on a trampoline instead of a wakeboard until you pointed it out. I think the water in the background confused me. I was also probably biased by reading the network's label before reaching my own judgment.
- jameshart 12y agoRight - impressive though this is, also the cat isn't black, the woman with the bananas might well not be a woman, the girl isn't swinging on a swing, the construction worker is probably not in the road, and the guitarist is tuning the guitar, not playing it. It's pretty much wrong -in detail- on every count. The weird thing is how at a glance, it seems pretty much correct. And how some people here are willing to look at that and think 'automated law enforcement is clearly imminently possible'.
- negamax 12y agoWow.. talk about high expectations. This is really well done. Identifying multiple concepts in an image and describe it to a good degree of accuracy. And this will get better further.
- jameshart 12y agoNo no! Don't misunderstand me! I agree: It's phenomenally impressive! It's also still completely useless at this level of accuracy. At this point, this software is as useful at describing photographs as a disinterested teenager who is busy trying to text. "This is a picture of my mom with some dude playing I dunno like tennis or something. Whatever." Which is really impressive! Seriously! But closing the gap to accuracy is really important, and it's a hard hard problem.
- dTal 12y agoAnd those are clearly plantains, not bananas!