MSE 491 Project
The first engineering technical elective course I took in my undergrad was MSE 491 - Application of Machine Learning in Mechatronic Systems. In this course, I learned some basic Python and some Machine Learning principles and applications. For my final project, I created a posture detection software using some body recognition libraries and algorithms.
The project aimed to improve posture while sitting at a desk using machine learning and negative reinforcement. The project resulted in the development of a posture detection software that uses webcam video input to monitor a person's posture in real-time. The software uses a machine learning algorithm to detect when a person is slouching or has poor posture. When this is detected, the software generates a buzz sound until the person corrects their posture. The idea behind this is to use negative reinforcement to encourage the desired behavior of maintaining good posture while sitting at a desk.
To develop the software, I used Python and Jupyter. I first trained the machine learning model using a dataset of images of people with good and poor posture. I then integrated the model into the software and tested it using webcam video input. Overall, the project was a success in that it was able to accurately detect poor posture and provide a means of encouraging improved posture through negative reinforcement. This has the potential to be a useful tool for people who spend long hours sitting at a desk and want to maintain good posture to avoid back pain and other health issues. In the future, I plan to continue refining the software and potentially integrating it into a more user-friendly interface, such as a mobile app. I believe this project has the potential to make a positive impact on people's health and well-being, and I am excited to see where it goes in the future.