Extraction Of Biological Features And Intelligent Sense

With development of technologies,such as big data,sensors,and deep learning,artificial intelligence will become a disruptive force of industries.Artificial intelligence in certain tasks,such as computer vision and speech recognition,has obtained outstanding achievements.But it is still easily confused by complex scenes or objects in complex scenes,which are very simple for humans.There still exist poor autonomy,poor human-computer interaction and other issues in the application of artificial intelligence into the daily life.To solve issues of poor autonomy and poor human-computer interaction,the paper put forward a concept of ’Intelligent Sense’.Core of the concept is that machine can perceive human information autonomously and environment can adapt to human spontaneously.The paper analyzed basic conceptions and compositions of the concept,and also expounded how to apply the concept to various industries.Based on the concept,autonomous perception of human information and spontaneous adaption to human without any input of human can be achieved through interactions among human,machine and environment.Autonomous perception of human information is an important part of ’Intelligent Sense’.With the goal of autonomous perception of human information,the paper studied how to acquire human information by camera instead of special sensors,and explored features that can characterize human states.By using cameras,non-contact acquisitions of ambulatory blood pressure and heart rate were realized.Speaker recognition was achieved through deep neural network.By non-linear method,features that can characterize human emotion and nervous system activity were obtained from neonatal cries.