UTILIZING A MACHINE LEARNING MODEL TO IDENTIFY ACTIVITIES AND DEVIATIONS FROM THE ACTIVITIES BY AN INDIVIDUAL

摘要

#$%^&*AU2020203106A120200528.pdf#####ABSTRACT A method, including receiving, by a device, configuration information associated with configuring an application for monitoring an individual, wherein the configuration information includes at least one of: information identifying physical characteristics of the individual, information identifying medications taken by the individual, personal information of the individual, or information associated with a caregiver of the individual, receiving, by the device, historical information associated with the individual, wherein the historical information includes at least one of: information associated with a health history of the individual, information associated with health histories of other individuals, information associated with activities of the individual, or information associated with activities of the other individuals, creating, by the device, a training set using the configuration information and the historical information, training, by the device and using the training set, a machine learning model to generate a trained machine learning model, receiving, by the device and via the application, monitored information associated with the individual from one or more client devices associated with the individual, the one or more client devices including at least one of an image sensor or an audio sensor, the monitored information including first monitored information including at least one of: a first video captured by the image sensor, or first audio captured by the audio sensor, and the monitored information including second monitored information representing information captured at a time subsequent to capture of the first monitored information and including at least one of: a second video captured by the image sensor, or second audio captured by the audio sensor, processing, by the device, the first monitored information, with the trained machine learning model, to identify one or more first activities of the individual, determining, by the device, a routine associated with the individual based on identifying the one or more first activities of the individual, processing, by the device, the second monitored information, with the trained machine learning model, to identify one or more second activities of the individual and one or more deviations from the routine by the individual, the one or more deviations determined based upon analyzing the second video or the second audio and analyzing the first video or the first audio, and performing, by the device, one or more actions based on identifying the one or more second activities of the individual and the one or more deviations, the one or more actions includingone or more of: causing a robot to provide medication to the individual based on a first deviation of the one or more deviations, or causing an autonomous emergency vehicle to traverse a route to the individual based on a second deviation of the one or more deviations.1/12 Cao E 0 1 00 , 0 0)a0 LP = c 0) 00 C)l ca aa 0 ca) 0 a) 00 cm 0 00 CDC .20 a)a .2 C ' Eca 0 02 a 0 ca) 'o E 0 l C _2 o I ca 0 0- U) ) a) 0 ~ E ca) 0~0 co 0 '91 L 3 0a a 0m