🎶 SongSmith | New remixes at your fingertips! 🎶
SongSmith is your one-stop-shop to create new remixes from your favourite songs! 🎼
State-of-the-art AI Machine Learning Neural Networks are used to generate remixes with similar styles🎨 to your best-loved songs🎶.
📚Discover! 👨🎤Inspire! 🎧Listen! SongSmith!
Inspiration 💭⚡
Ever listen to your favourite artists🧑🎤 and songs⏭️ and think "Damn, I wish there was similar music to this?" We have this exact feeling, which is why we've developed SongSmith for the Music loving people!🧑🤝🧑
How we built it 🏗️
- SongSmith⚒️ was built using the latest and greatest technology!! 💻
- Our music generative neural network was developed using a state-of-the-art architecture🏢 called Attention Mechanism Networks.
Tech Stack 🔨
- AI Model: Tensorflow, Keras, Google Colab
- BackEnd: Express.js, Flask to run our microservices, and inference servers.
- FrontEnd: Developed using React.js, and Bootstrap; to allow for a quick MVP development cycle🔃.
- Storage: MongoDB, Firebase, Firestore
- Moral Support: Coffee☕, Bubble Tea🥤 , Pizza🍕 and Passion💜 for AI <3
Challenges we ran into 🧱🤔
- Converting the output from the neural network to a playable format in the browser
- Allowing for CORS interaction between the frontend and our microservices
What we learned🏫
- Bleeding💉 Edge Generative Neural Network for Music🎙️Production
- Different Python🐍 and Node.js Music Production Related Libraries📚
What's next for SongSmith ➡️
- Offering wider genres of music to attract a wider range of music-loving users!🧑🎤
- Refining our Neural Network to generate more high-quality REMIXES!
*See Github ReadMe for Video Link
Built With
- bootstrap
- express.js
- flask
- javascript
- keras
- node.js
- python
- react
- tensorflow
Log in or sign up for Devpost to join the conversation.