@Singularitarian what about the recent image recognition results
-
-
Replying to @Shedletsky
@Shedletsky@Singularitarian It's very impressive, but news reports saying "OMG better-than-human" are very misleading1 reply 0 retweets 0 likes -
Replying to @michael_nielsen
@michael_nielsen@Singularitarian I saw the results where they had a program that could recognize different breeds of dogs.1 reply 0 retweets 1 like -
Replying to @Shedletsky
@michael_nielsen@Singularitarian Most people can probably only name 20-30 different breeds, so it's probably super-human in that regard3 replies 0 retweets 1 like -
Replying to @Shedletsky
@michael_nielsen@Singularitarian although by my logic a dictionary is super-human in that regard... however the pace of progress is v. fast3 replies 0 retweets 1 like -
Replying to @Shedletsky
@Shedletsky@Singularitarian @karpthy has a nice blog post on the process: http://karpathy.github.io/2014/09/02/what-i-learned-from-competing-against-a-convnet-on-imagenet/ …2 replies 0 retweets 0 likes -
Replying to @michael_nielsen
@michael_nielsen Nice post. I think performance is super-human at this point. The winning teams are at 3.5% as of 2015 vs 6% for this guy.2 replies 0 retweets 0 likes -
Replying to @Shedletsky
@Shedletsky For ImageNet top-5 criterion. This doesn't correspond to any particularly interesting human criterion.1 reply 0 retweets 0 likes -
Replying to @michael_nielsen
@michael_nielsen but it is better than the author was able to do with the same criterion. Whether it is interesting or not is besides the pt1 reply 0 retweets 0 likes
@Shedletsky The claim "better-than-human vision" is simply incorrect, for definitions of "vision" that humans care about.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.