Our program detects negative emotions through audio, and releases appropriate scents to improve your mood. Over time it will personalize which scents are most effective for you. The effect of the scents are further enhanced by subconsciously conditioning you to respond positively to them.

Audio from a microphone in the home(Google Home, Amazon Echo, etc.) is constantly analyzed for emotion. When the user is angry, anger-reducing scents are released. When the user is unhappy, mood-boosting scents are released. User Experience: 2-week Calibration Period to determine best scent for user: - Starts with a default set of scents for each emotion - Happy scents: Lemon, Lavender, and Jasmine - Calming scents: Rosemary, Cinnamon, and Peppermint - By analyzing change in emotion values before and after the release of scent, our algorithm will determine which scent is most effective for each emotion for the user

Post-Calibration (Classical Conditioning): - After the algorithm down-selects to one scent for happiness and one scent for calming, those scents are used to condition the desired response. - Example: Lemon is found to be most effective at making user X happy. When user X is unhappy, lemon is sprayed, and natural effect of the scent boosts their mood. After this happens many times, user X’s mind makes a connection between lemon and the mood boost that follows. - Smelling Lemon —> Happiness increase - The conditioned response enhances the natural mood enhancing effect of the lemon

Built With

Share this project: