Re:The ImageNet dataset transformed AI research
by PeterisP on Thursday July 27, 2017 @12:10AM
(#14862646)
99.9% is an extremely high bar. People aren't machines, people make errors. If you asked them to tell whether a square is black or white, you would get less than 99.9%.
[ Parent
| on HN ]
Re:The ImageNet dataset transformed AI research
by ouid on Thursday July 27, 2017 @12:55AM
(#14862811)
99.9% is a very low bar for image recognition. if one out of every thousand times you saw a car you misclassified it as a chair, you would be dead.
Re:The ImageNet dataset transformed AI research
by mturmon on Thursday July 27, 2017 @01:24AM
(#14862884)
It sounds like you may not have experience in this domain? You're benchmarking an active, time-based task (a person "seeing a car" - implicitly allowing eye saccade, multiple images, context, depth) versus a static task ("classify this single image based on no context"). They are not comparable.
Re:The ImageNet dataset transformed AI research
by ouid on Thursday July 27, 2017 @02:51AM
(#14863173)
have you actually tried the imagenet task yourself?
The reason imagenet is popular is because it has allowed google to claim that they can beat human performance on image recognition by artifically lowering human performance on image recognition.