MusFinder is an intelligent music recommendation system designed to
integrate with existing music applications, enhancing users' listening
experience by providing more personalized and context-aware music
suggestions.
Low-Fidelity and Working Prototype
Demo Video
Highlights of Evaluation Result
Integrated a Large Language Model (LLM) API and built a user-friendly
interface compatible with existing regular music applications.
Photo/Keyword Input: 100% task success; Users found it intuitive and
matched to mood.
Context-Aware Recommendations: 80% saw appropriate music adjustments
based on location.
Real-Time Feedback: 80% reported better recommendations after feedback
input.
User Feedback: Praised interface ease and music match quality; suggested
faster feedback response.
Behavioral Observations: Users explored freely with minimal guidance;
slight hesitation with image uploads.
