Palette of Emotion

Inspiration

Have you ever been overwhelmed by strong emotions with no one to share them with? In this concrete jungle, we all face intense emotions daily, yet finding someone who truly understands us is a challenge. This loneliness can lead to depression, poor sleep quality, impaired executive function, and accelerated cognitive decline, according to the American Psychological Association. Psychologists have found that visualization can greatly improve mood and reduce anxiety Inspired by art therapists who use visual media to help clients express their emotions, in this hackathon, we have created a website that generates a personalized emotional palette, incorporating both historical and modern interpretations, to visually represent user's feelings.

What it does

Our project leverages AI to bridge users' emotions with art, offering a unique, interactive experience. By inputting their feelings, users initiate an analysis process using the OpenAI API, which then identifies artworks from the MET Museum that resonate with their emotional state. Enhancing this personalized journey, our system employs Dall-E to create custom artworks inspired by both the user's emotions and the selected MET Museum pieces. This innovative approach not only fosters a deeper connection between individuals and art but also supports emotional well-being through creative expression and reflection.

How we built it

Our approach integrates emotional well-being with technology in three concise steps. First, we collect emotional inputs from users, fostering self-awareness and mental health benefits. Next, we analyze these emotions using the OpenAI API to find matching artworks in the MET Museum's collection, creating a personal connection to art as a form of emotional support. Finally, we employ Dall-E to generate unique artworks inspired by both the user's emotions and the chosen MET Museum pieces, merging art history with AI creativity for a therapeutic experience.

Challenges we ran into

Leveraging the OpenAI API for artwork analysis presented challenges due to its lack of training in this specific domain. Fortunately, the fame of MET Museum artworks allowed for more effective analysis.
The total runtime of approximately one minute exceeded our desired speed, potentially affecting user experience. Optimizations through improved computing resources and a refined local database are anticipated to enhance performance.