Inspiration
Upon examining the initial challenges, the concept of "BrainBrush" caught our attention. Intrigued by the idea, we decided to enhance it by incorporating an AI image generator. Motivated by the recent surge in AI art, we aimed to integrate this technology with a Brain-Computer Interface (BCI). The notion of free association also influenced our program, drawing inspiration from the concept of allowing subjects to express their thoughts without censorship. In essence, our program acts as a form of free association, projecting the subject's emotions onto AI-generated art without constraints.
What it does
Our program utilizes signals from Muse 2's EEG, translating Delta, Theta, Alpha, and Beta waves into emotions. We leverage an online resource (https://github.com/alexandrebarachant/muse-lsl) to fetch the data and then map these emotions to associated words and generate AI images using tools like stable diffusion.
How we built it
Utilizing muselsl (https://github.com/alexandrebarachant/muse-lsl), we establish a connection to the Muse 2 device to extract EEG data. Leveraging these resources and our background knowledge, our program discerns whether the user is experiencing stress and quantifies it into a stress level. We've curated various predetermined terms that represent stress levels across different scales. Using a combination of these terms in a sentence, our system prompts the user with an image generated by the Stable Diffusion API.
Challenges we ran into
Time constraints hindered our plan to implement a stress detection system. The original idea involved pausing the program or finding ways to calm the subject if excessive stress was detected.
Accomplishments that we're proud of
Despite being the first-time participants for many team members, we successfully completed our program during the hackathon. Efficient communication and collaborative brainstorming were key strengths.
What we learned
We gained insights into neurotech and BCIs, recognizing their potential to significantly impact the world. Our team learned how to work cohesively and acquired knowledge about obtaining and interpreting data from neurotech devices for use in various programs.
What's next for Untitled
Given more time, our program could undergo refinement and contribute to research, recreation, and rehabilitation. In research, as neurotech advances, our program could potentially interpret more specific aspects of brain activity, such as thoughts and ideas, allowing us to translate these into images. For rehabilitation, our program provides an alternative means of expression for individuals with difficulties, such as those with ASD or social anxiety, enabling communication through art and generated images. Additionally, the program could be applied in psychotherapeutic free association, assisting patients in expressing thoughts and emotions without constraints. Lastly, the program has recreational potential, offering artists and creators the ability to envision and bring their ideas to life.
Log in or sign up for Devpost to join the conversation.