Selin ’27 is designing a device to detect and respond to misophonia-triggering sounds using a self-taught algorithm and white noise system, combining personal motivation, persistence, and coding to improve quality of life for those affected by the disorder.
This summer I started creating a device to address the challenges of the presence of misophonia triggers in daily life. Misophonia is a disorder where certain repetitive sounds can trigger negative reactions that can impact a person’s quality of life significantly. This disorder can cause intense anxiety, rage or even severe isolation, greatly influencing one’s mental and emotional health. I first observed the presence and implications of misophonia in my family, and when I decided to research more about the condition, I became very passionate about finding a way to reduce the negative impacts misophonia has on the lives of my loved ones and my community. So, I made a plan to develop a noise response system for misophonia triggers.
I found that white noise is a great way of covering misophonia triggering sounds and reducing the change or severity of reactions. Based on the pilot data of a study showcasing the discomfort ratings participants with some level of misophonia gave to different sounds, I chose the eight sounds with the heights discomfort ratings from the study as the major triggering sounds for my project. I planned on using a training model to create an algorithm that would take in a sound from its environment, determine whether it is a misophonia triggering sound or not, and start playing white noise if it recognized the sounds as one of the eight identified triggering sounds.
When I think of how I have been showing up for this project, I think of the fact that at the start of this project, I did not have a lot of coding knowledge, especially about training models. I definitely came across many instances where I had no clue how to approach a problem or didn’t have enough knowledge about the coding language or context about the step I was at. This was pretty demotivating, especially at the start, and it became a challenge in regards to wanting to keep showing up. But that is probably what should be anticipated with every project, so I kept on researching, working by trial and error, debugging the algorithm at every step, and pushing forward through every challenge.
Programming an algorithm and turning it into a device is an extensive process, so creating a schedule to follow and establishing small goals to reach throughout the project has been really important in keeping me on task and making sure I put in the effort that is needed to have a successful outcome. It is easy to get lost in the many steps of producing a product so I have found that using structured checklists for each step of the way has been a great way of using my organizational skills to stay on track. At first I started to code my algorithm on google colab, now I am transferring my training model code onto a Raspberry Pi Compute Model and continuing to code the final steps of my algorithm there. I’m aiming to have a functional device or app developed by the end of the summer.
