Inspiration
Public speaking and performing is a common phobia around the world. While there are techniques like 4-7-8 breathing method for managing the anxiety, they all share a common flaw: they lack a real-time feedback loop. To address this gap, we created a device that moves past basic relaxation techniques and offers a visual, real-time feedback of a speaker’s actual confidence.
What it does
Pulse & Podium is a wearable biofeedback system that quantifies public speaking confidence. By monitoring Skin Conductance (GSR), Skin Temperature, and Heat Rate, our device detects the subtle change on body. Instead of guessing if a breathing technique is working, our algorithm provides a real-time "Confidence Score." This allows speakers to see exactly how their body responds to the pressure of the podium, turning abstract feelings into manageable data.
How we built it
We start with designing the Arduino components, such as the heart rate monitor, the GCR and the body temperature sensor. At first we tried Figma. We worked on creating a custom dataset with our own voice, taking 225 recordings. However, this lead to overfitting and we discovered Google AI Studio to make a helpful generalized assistant that offers helpful feedback.
Challenges we ran into
At first the AI from Figma was not applying the arduino components correctly. As a result we switched to Google AI studio.
Accomplishments that we're proud of
Creating a custom data set for the first round of AI training that was created by the team. We are also proud of sending real time data to the site, and the interactive feel of the website.
What we learned
We cannot have serial plotter open while also running the site, as it can't send data to both places. We also learned more programming with arduino, and how to use AI that is embedded in out projects.
What's next for Pulse & Podium
Training the model with more data sets, and ensuring accurate data is sent to it, and the model can interpret the values better.
Log in or sign up for Devpost to join the conversation.