Knock! Knock! Who’s there? – Artificial Neural Network and Deep Learning Modeling
School Name
Governor's School for Science and Mathematics
Grade Level
12th Grade
Presentation Topic
Engineering
Presentation Type
Mentored
Written Paper Award
3rd Place
Abstract
Analyzing vibrational patterns, such as knocking, for identification can be used to help with safety and security. Applications of the identification process can be used in the Internet of Things industry and smart technology, such as smart homes. The objective of this research project was to use artificial neural networks to measure the vibrations of a door caused by a person knocking, and use that data to identify the knocker. One hundred sets of knock vibration data were collected from five test subjects and compared to ensure that this project was actually feasible. The experiment showed that although there was a lot of variation there was a distinct pattern in the overall knocks. A second program was written using artificial neural network technology to train the computer to learn the knock vibration patterns and to use this data to identify the person knocking. The results were able to find that knock vibrations can be used to identify a person, and that can lead to the prospect of using more sensitive vibrations to identify and detect humans.
Recommended Citation
Chaubey, Ridhi, "Knock! Knock! Who’s there? – Artificial Neural Network and Deep Learning Modeling" (2018). South Carolina Junior Academy of Science. 45.
https://scholarexchange.furman.edu/scjas/2018/all/45
Location
Neville 109
Start Date
4-14-2018 8:45 AM
Presentation Format
Oral and Written
Knock! Knock! Who’s there? – Artificial Neural Network and Deep Learning Modeling
Neville 109
Analyzing vibrational patterns, such as knocking, for identification can be used to help with safety and security. Applications of the identification process can be used in the Internet of Things industry and smart technology, such as smart homes. The objective of this research project was to use artificial neural networks to measure the vibrations of a door caused by a person knocking, and use that data to identify the knocker. One hundred sets of knock vibration data were collected from five test subjects and compared to ensure that this project was actually feasible. The experiment showed that although there was a lot of variation there was a distinct pattern in the overall knocks. A second program was written using artificial neural network technology to train the computer to learn the knock vibration patterns and to use this data to identify the person knocking. The results were able to find that knock vibrations can be used to identify a person, and that can lead to the prospect of using more sensitive vibrations to identify and detect humans.
