BackTrack

Inspiration

As kids, we would constantly be losing our belongings, whether it be our backpacks at a playground, water bottles at school, or jackets after practice. It's never fun to lose something. Keeping this in mind, we were inspired to create BackTrack, a simple and reliable way to find your lost items.

What it does

BackTrack is a lost-and-found platform that connects people who’ve lost something to those who've found it.

When someone loses an item, they can upload a photo or description of it to BackTrack. Similarly, when someone finds an item, they can do the same. Our system then analyzes and compares the images, identifying potential matches based on visual similarity.

How we built it

We built BackTrack by developing a Python-based system that converts both image and text inputs into vector representations. We then compare these vectors to efficiently measure similarity between lost and found items. The resulting vectors are stored securely in AWS S3, which enables fast retrieval and comparison across large datasets of items. We used Cloudflare to manage our domain routing and network configuration, allowing for secure and efficient flow into our backend. Through Cloudflare, we set up HTTPS with SSL encryption to provide a reliable, scalable, and secure connection for our users.

Challenges we ran into

One of the biggest challenges we faced was generating and comparing vector data, as ensuring accuracy and consistency between image and text inputs took a lot of testing and tuning. We also ran into difficulties converting image data into a usable format, specifically transforming photos into Base64 data that could be processed and stored efficiently. Managing the integration between our Python backend, AWS S3 storage, and Cloudflare service also was challenging, especially in keeping the data retrieval fast and reliable.

Accomplishments that we're proud of

We're proud of being able to build a working prototype which can accurately extract and compare our vector data. We also are proud that we could use AWS, Cloudflare, etc. to do this and that we figured out how to configure such large scale technology for our own uses.

What we learned

We learned a lot about vector embeddings and how vectors worked in general, but also how we can use services like AWS's S3 to help us with our programs. Beyond the technical side, this project helped us learn more about backend infrastructure and how different technologies can work together.

What's next for Backtrack

We plan to add real time notifications, a lost to found texting feature, and locational data to help ease user experience. We would also love to integrate BackTrack into other platforms like MyUW to more centralize the process and expand user usage.