Bard GO! | Devpost

Inspiration

For some of us its while we are studying, or while we are conversing with friends and having fun. We wake up in the morning and we think, "I wish I had a soundtrack for my life." Hence we came up with Bard GO!

What it does

Through a webcam, our app analyzes the surroundings the user is currently in, and produces background music accordingly.

How we built it

To build our app we used PyTorch, Hugging Face, and the Spotify API. We trained a model that converts images to emotion vectors. We used an algorithm to play the songs to give it the best experience possible depending on the mood. We used the Spotify API to get the song that the user wants to be played, get info on the songs to find features for the algorithm, and for playback.

Challenges we ran into

The main challenge we faced was that we had many drafts of ideas and we had to constantly switch between plans in order to achieve our goals. Our plans were often very detailed and we often had to draw out diagrams to visualize them. We put a lot of thought into the algorithms we used such as the song switching algorithm Our original design involved creating an iOS client that would seamlessly integrate with the Spotify API, a Flask API server, and a CockroachDB or JSON style database, but we encountered numerous obstacles along the way, ultimately forcing us to abandon this approach. Instead, we pivoted to using the tools listed in the section above as the backbone of our project. As the deadline loomed, last-minute changes became inevitable, reflecting the dynamic nature of our development process.

Accomplishments that we're proud of

We put in a lot of effort and worked diligently to let our idea come into life.
We are happy with the logo we designed
We are proud of how much we learned to build this project
We we collaborated well on this project
We showcased our ability to chase directions when something challenges arose which is an important skill to have
We are proud of our one the spot creativity and time management
We are especially proud of our determination and our willingness to finish this project despite all the challenges that arose

What we learned

Our journey was characterized by a focus on the process rather than the end result. While several of the initial technologies we explored didn't ultimately find a place in our final product, we delved deeply into these tools before making the decision to pivot our plans.