| dc.description.abstract | Crying is a major form of expression that can be issued or displayed by a baby.
Through crying, babies communicate their needs to parents such as hunger, fatigue, pain
and other unpleasant body conditions, as well as to fulfill their desire to be cared for. Often
housewives are tired of rocking their children to sleep, to help ease the mother's task it is
necessary to have a tool that can cheer up automatically when the child is crying. One of
them is an automatic swing tool. Crying which is a sound wave can be captured by a sound
detector, in this case we call Voice Recognition named Elechouse V3 with an Arduino Nano
slave and a Load Cell with HX711 as a baby load detector, by utilizing Wemos D1 ESP
8266 as the main controller to control the Stepper motor which is also connecting the
system with an android device via the Blynk application. In this study, a test was performed
on the accuracy of the detection of a baby's cry and not a baby's cry. From the test results,
it was found that the accuracy of crying detection according to the train sample was 100%,
with an average detection duration of 3.7 seconds and a median detection duration of the
2nd and 6th seconds each detected 3 times. And the accuracy of non-crying sounds that are
not considered as a baby's cry is 100%. | en_US |
