LAUSR

Human-inspired computing is the intelligence computing model enlightened by human brain intelligence and biological processes. It is not only the inheritance and development of artificial intelligence, but also from a new point to understand and grasp the intelligent intrinsic. With the recent advances in cognitive science, we now have the ability to mine more details of the human brain to obtain valuable insight about what is happening in the brain. Enlightened from these new achievements in the brain cognition, scientists have proposed many new theories and algorithms of human-inspired computing, which have been applied widely in many fields, such as artificial intelligence, machine learning and data mining. Human-inspired computing is becoming a significant force to promote science and technology progress. The goal of this special issue is to call for a coordinated effort to understand the opportunities and challenges emerging in human-inspired computing, identify key tasks and evaluate the state of the art, showcase innovative methodologies and ideas, as well as introduce real systems or applications, and discuss future directions. Intelligence computing plays an important role in public security, entertainment, healthcare, etc. We solicit manuscripts in the important fields of intelligence computing that explore the synergy of human-inspired computing and deep learning techniques. In the end, six papers are accepted to this special section. The first paper, entitled “Why and when can deep-but not shallow-networks avoid the curse of dimensionality: a review” from Tomaso Poggio et al., reviewed and extended an emerging body of theoretical results on deep learning. By comparing shallow (one-hidden layer) networks with deep convolutional networks, the authors investigated the conditions under which deep networks can be exponentially better than shallow learning. Conclusion was drawn that deep networks have the theoretical guarantee to avoid the curse of dimensionality for compositional functions, such as locality of hierarchy and compositional input-output mapping. The second paper, entitled “Imitating the brain with neurocomputer – a “new” way towards artificial general intelligence” from Tie-Jun Huang proposed a practical approach to achieve artificial general intelligence by imitating the connection structure of brain to build up neurocomputer, based