HIGHLIGHTS
SUMMARY
(A) Synset Animal - Target - (B) Synset Animal - Distractor - (C) Synset Artifact - Target - (D) Synset Artifact - Distractor - Human categorization of an animal can be performed with high accuracy (generally over 80% correct), very quickly, and is robust to geometric transformations. The linear part of the processing is performed by a convolutional operator, hence the name convolutional neural_network (CNN) for this class of architecture. Unlike artificial neural_networks, which can easily compare these 1000 possibilities simultaneously, one can instead use a subset of behaviorally relevant labels to make the task more relevant to humans. Using a set . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.