-
Initial Login Page
-
Twilio Message Authentication
-
Chat Interface that uses gpt-3.5 LLM to interact
-
Output Generated. Play Music button will generate the audio (voice + music)
Inspiration
As music lovers, we often find ourselves wondering what genre a particular song belongs to. We also enjoy writing poetry and playing music. Combining our interests, we came up with the idea of Audio Poet, a web app that generates lyrics using ChatGPT, identifies the genre of the lyrics using a machine learning model, and maps it to appropriate music, creating a unique audio experience.
What it does
Audio Poet is a web app that generates lyrics using ChatGPT, identifies the genre of the lyrics using a machine learning model, and maps it to appropriate music. The app then plays the music along with a bot voice that reads the generated lyrics, creating a unique audio experience.
How we built it
We built Audio Poet using a combination of technologies. For generating the lyrics, we used ChatGPT, a large language model trained by OpenAI. Backend is written in Python Flask and JavaScript for buildng a simple interface. We also used a machine learning model to identify the genre of the lyrics. To map the genre to appropriate music, we used a music library and an algorithm that selects the best match based on the genre. To play the music along with the bot voice, we used a text-to-speech engine and a music player. Stored the generated lyrics in NoSql DB (MongoDB)
Challenges we ran into
One of the biggest challenges we faced was training the machine learning model to accurately identify the genre of the lyrics. We also had to find a way to map the genre to appropriate music in real-time, which required a lot of experimentation and fine-tuning.
Accomplishments that we're proud of
We're proud of creating a web app that combines cutting-edge technologies like ChatGPT and machine learning to create a unique audio experience. We're also proud of the algorithm we developed to map the genre to appropriate music in real-time.
What we learned
Through this project, we learned a lot about natural language processing, machine learning, and music theory. We also learned about the challenges of integrating different technologies to create a seamless user experience.
What's next for Audio Poet
In the future, we plan to improve the accuracy of the machine learning model and add more genres and music options to the app. We also plan to add a feature that allows users to upload their own lyrics and map them to appropriate music. Additionally, we plan to improve the bot voice to make it sound more realistic.
Log in or sign up for Devpost to join the conversation.