2025 Summer Undergraduate Research Program (SURP) symposium
Location
Dr. Ken Budke Family Auditorium, Schindler Education Center, University of Nothern Iowa
Presentation Type
Open Access Poster Presentation
Document Type
poster
Abstract
Non-communicable gum disease affects about 42%[4] of adults aged 30 years and older. Developing a robust model to detect brushing techniques on a smartwatch can help reinforce good brushing habits and prevent gum disease.
The likelihood of someone owning a smart watch compared to a smart toothbrush is staggering. There were around 66,051[3] smart toothbrushes sold in 2022, with smartwatches shipping about 150,000,000[2] units. It is important to have an alternative on a smartwatch, because it will reach far more users.
Data privacy has become a large issue worldwide. With a pre-trained model that is systematically fine-tuned by the program, users can rest easy knowing that their data never leaves their device, but their experience can still improve.
Start Date
1-8-2025 11:00 AM
End Date
1-8-2025 1:30 PM
Event Host
Summer Undergraduate Research Program, University of Northern Iowa
Faculty Advisor
Dheryta Jaisinghani
Department
Department of Computer Science
Copyright
©2025 Anthony Teal and Dheryta Jaisinghani
File Format
application/pdf
Recommended Citation
Teal, Anthony and Jaisinghani, Dheryta, "Tooth Brushing Technique Detection With Deep Learning on the Edge" (2025). Summer Undergraduate Research Program (SURP) Symposium. 1.
https://scholarworks.uni.edu/surp/2025/all/1
Tooth Brushing Technique Detection With Deep Learning on the Edge
Dr. Ken Budke Family Auditorium, Schindler Education Center, University of Nothern Iowa
Non-communicable gum disease affects about 42%[4] of adults aged 30 years and older. Developing a robust model to detect brushing techniques on a smartwatch can help reinforce good brushing habits and prevent gum disease.
The likelihood of someone owning a smart watch compared to a smart toothbrush is staggering. There were around 66,051[3] smart toothbrushes sold in 2022, with smartwatches shipping about 150,000,000[2] units. It is important to have an alternative on a smartwatch, because it will reach far more users.
Data privacy has become a large issue worldwide. With a pre-trained model that is systematically fine-tuned by the program, users can rest easy knowing that their data never leaves their device, but their experience can still improve.
