tune.ai

Imagine typing in "sultry like Lana Del Rey," or "in the summery feel of The Beach Boys," and having a customized track created to reflect that specific ambiance.

Comment

Inspiration

The journey of blending technological innovation with creative artistry has always been an exciting avenue to explore. Our inspiration for this project sprouted from a fascination for music, an appreciation of diverse artists, and a curiosity to witness the marvels of AI and machine learning firsthand. We wanted to devise a system that enables users to generate unique music for all their needs while respecting their preferred vibe and artist inspiration. But most of all, we were tired of listening to the same songs by the Strokes!

What it Does

Our platform, tune.ai, is an AI-driven music generator capable of creating music based on a specific vibe and/or an artist's style. It's a perfect tool for musicians seeking inspiration, businesses developing unique tones for their brand, podcasters, filmmakers, or anyone who loves music.

How We Built It

We rigorously expanded on Meta Audiocraft's MusicGen model to generate tunes which align with the given vibe. We added an option to choose an artist's style through UM-GPT's API. The user-friendly web interface was created using HTML/CSS/JS for front-end design and Flask for back-end functionality.

Challenges We Ran Into

While developing tune.ai, integrating two different AI models—Meta Audiocraft's MusicGen and UM-GPT— and communicating between them presented some challenges such as formatting responses, debugging and optimizing code, and finding a balance between runtime and audio length.

Accomplishments That We're Proud Of

Successfully creating a musical platform that encapsulates specified vibes and employs the nuances of individual music styles is a tremendous feat we're genuinely proud of. Our system is robust, user-friendly, and a truly unique approach to AI-driven music generation, because we allow the user to have unlimited choice in the vibe and artist that they want their tune to resemble as opposed to having limited "check-in-the-box" options like most AI-driven music generators.

What We Learned

Developing tune.ai was an enriching experience; we learned to integrate two complex AI models, design a responsive web interface, handle unanticipated issues, and above all, we gained deeper insights into bridging the gap between technology and creative arts.

What's Next for Tune.AI

We envision continually evolving our platform. With the advancement of artificial intelligence, we plan to enhance its capability to incorporate further customization features, add a wider spectrum of musical genres, and improve the audio quality of the generated music. Our goal is to propel tune.ai to become an essential tool for diverse musical needs, providing users with unparalleled music generation experiences.

Share this project:

Updates