All Roads Lead to Matcha

Inspiration

We were inspired by @Spu7Nix's amazing algorithm that turns any image into Obama and decided to take a more hilarious and even more useless take on it!

What it does

Our project takes any input image and finds a hilarious chain of Wikipedia page connections, reminiscent of "6 connections". It transforms and rearranges the pixels in the input image to resemble the next image in the path, while narrating a humorous song based on the connections. Additionally, it provides a conversation guide for the user to hijack any interaction and steer it to display their performativity.

How we built it

We used Gemini (2.5 flash pro) will find the Wikipedia connections between the uploaded image The application shows the chain of connections between images (Example, Obama to USA, USA to China, all the way until the desired performative item). It also provides the Wikipedia links to each page, if you would like to do some more performative reading!

Challenges we ran into

We ran into trouble implementing the audio features in enabling the program to rap the topic-generated lyrics, but were able to overcome the issue using Gemini API.

Accomplishments that we're proud of

We combined several features and maximized our use of Gemini API. We also challenged ourselves to further diversify and improve our project throughout the day, which definitely pushed our project much further than we had initially visualized.

What we learned

We learned a lot about implementing Gemini API in our projects and balancing vibecoding with structured development.

What's next for All Roads Lead To Matcha

As performative males become too mainstream, we will look to introduce 'elo' and ranking to users, and suggest a broader range of niche categories of performativity.