Inspiration
We were inspired by the prospects of computer vision and being able to bridge that gap with the sounds of music. If we could connect our ideas together, we could convert vision into vibe.
What it does
We using hand gestures, we can take an image and send that to convert our image to a text description which is mapped to a select few genres using the OpenAI API. Then, the program sends those genres to Spotify API and get the best recommendations that Spotify has to offer us.
How we built it
Gesture-activated smart webcam captures your scene With the use of AI and LLMs, we describe the image in text form and convert this to be mapped using OpenAI's API. We then give a curated set of the tracks which the Spotify API finds, it is the perfect music match for your vibe
Challenges we ran into
We weren't able to train our model from scratch like we had planned and so we switched routes after this hurdle.
Accomplishments that we're proud of
The computer vision was a major hurdle we faced but we were able to get a gesture to take a picture when said gesture is made. We were also proud of how we leveraged the APIs used to intertwine and complete the
What we learned
We learned about computer vision and gestures. We then were learning about how to make our own CNN and uses datasets to make this model
What's next for Vybe Fynd3r
We hope to make this idea into a wearable device which uses it's own model which is train to convert images-to-genre descriptions and continue improving it.
Built With
- css
- mediapip
- openai
- opencv
- python
- scikit-learn
- spotifyapi
Log in or sign up for Devpost to join the conversation.