Inspiration
Our modern world is saturated with invisible stress. People often internalize anxiety, either because they lack the language to express their feelings or the awareness to recognize them. An opportunity was identified to use music—one of the most powerful tools for influencing emotion—in a more intelligent and personalized way. Existing music recommendation systems rely on past behavior, but our feelings are instantaneous and dynamic. WAVES was born from this realization. The goal was to create a system that responds to how you feel in this exact moment. By using brainwaves as the core input, WAVES bypasses subjective self-reporting to create the most honest recommendation algorithm possible. This is especially powerful for individuals who may struggle with verbal communication, offering a direct pathway to emotional support and regulation through sound.
What it does
WAVES is a Brain-Computer Interface (BCI) that functions as a real-time, adaptive music recommendation engine. It creates a closed-loop system to help regulate and improve a user's mental state through sound. Wear the EEG Headset The user puts on a non-invasive Emotiv headset that reads their brainwave signals. Live Emotion Analysis The EMOTIV platform processes these signals in real-time to infer the user's emotional state, identifying key metrics like stress, focus, and relaxation levels. Adaptive Music Pairing This is the core of WAVES. The system intelligently selects and plays music from a curated library(alpha waves, beta waves, gamma waves) of functional audio designed to positively influence the user's current state. If it detects signs of stress, it might play calming ambient music known to entrain alpha waves. If it detects a lack of focus, it could switch to a beta-wave-associated track to gently nudge the user's mind toward concentration. The music adapts live, creating a seamless, responsive soundscape tailored to your needs.
How we built it
Brainwave Data Acquisition: Brainwave data is captured using an Emotiv EEG headset from five key channels (AF3, T7, Pz, AF4, T8), which cover the pre-frontal, temporal, and parietal lobes. Backend & Signal Processing: The backend is built in Python. The platform utilizes the MNE-Python library to clean the raw EEG data and perform Fourier transforms to extract the power spectral density (PSD) across different frequency bands (Alpha, Beta, Gamma, etc.). Emotion Classification at EmotivBCI app. Adaptive Recommendation Engine: The classified emotional state is fed into a custom recommendation engine. This engine queries a purpose-built library of functional music, with tracks tagged by their psychoacoustic properties. The engine's goal is to select the optimal track to mitigate negative states (like stress) or enhance positive ones (like focus).
Challenges we ran into
The primary technical challenge in developing WAVES was creating a reliable mapping between complex EEG signals and the music library. And we weren’t able to get a satisfactory result from the EEG Data An emotional state isn't a simple "on/off" switch, so the engine had to be designed to handle nuanced transitions and avoid jarring changes in the music. Filtering the noisy EEG signal in real-time to get a stable input for the model was also a significant hurdle that required careful engineering.
What we learned
The development of WAVES highlighted how to process complex biological signals and, more importantly, how to translate that data into a user-centric experience. It provided deep insights into the power of functional music as a tool for mental and emotional regulation.
What's next for WAVES
Right now, WAVES is your brain's personal DJ, picking the perfect track for your mood. But we're about to give it a promotion.
This will allow an individual to enjoy a modified environment, offering a powerful therapeutic tool for meditation and for users with ADHD and other neurodievergence. Expanding on this, WAVES will enable individuals groups to create musical remixes from their collective feelings. This provides a revolutionary non-verbal language for individuals with Autism or Dyslexia, fostering a new medium for empathy and collaboration. Refining our emotion-detection model to power this engine will make WAVES an essential tool for both personalized therapy and inclusive communication. In a word, tranform the way people connect and feel each other by tranlating brainwaves (specifically EGG signals) to soundwaves & music remix - a common language we all share, understand and appreciate.
Log in or sign up for Devpost to join the conversation.