Vibe Check

Inspiration

Spotify is not good. While Spotify’s algorithm tries to match our moods, it often misses the nuance of what we’re actually feeling. Music should understand the listener better — not just through skips and likes, but through genuine emotional context.

What it does

Vibe Check analyzes your emotional state either by reading a picture of your face or by interpreting a written description of your mood. Based on that analysis, it generates a personalized Spotify playlist designed to match or improve your emotional state — whether you need motivation, comfort, focus, or calm.

How we built it

We combined computer vision and natural language processing to interpret emotions from both images and text.

Used a pre-trained emotion recognition model for facial analysis.
Used a sentiment and tone analyzer for text-based input.
Connected to the Gemini API to generate playlists tailored to the detected mood.
Built in JavaScript, integrated APIs, and developed a simple frontend where users can describe their emotions.

Challenges we ran into

Mapping emotions based on the users descriptions and creating a playlist based on the description.
Managing Gemini API limitations and authentication tokens.
Claude API was not free, thus we had to alter all of our code to fit into Gemini API.

Accomplishments that we're proud of

Finishing the project, although we did have to scratch a good amount because of the API problems.

What we learned

We learned how to merge emotion analysis with real-world applications using APIs. We also discovered how powerful emotional data can be in creating more human-centered digital experiences, when handled ethically and responsibly.