Lost&Found

Inspiration

While trying to decide which muffin to purchase at Brown University's Blue Room, we overhead a woman asking the barista, "Has anyone seen my water bottle?" Losing things on campus is common, but there is no system for searching for an item that reaches the whole campus community. For example, Facebook groups often only include a specific class year; department mailing lists don't always include students and staff from other departments who may have been using the space they lost their item in.

What it does

Left a hoodie in your physics lecture classroom? Found a set of headphones on a dining hall table? Never fear, Lost And Found (LAF) is here! LAF is a web application that provides users with a method to report items that have gone missing or have been found.

How we built it

Python 3.7 Flask HTML/CSS and JS →Front end site development Google Cloud Vision API →Using pre-trained models to extract text labels from uploaded object images Google Maps API →Customize pins →Navigating the map dynamically Secure mailing system →Using middleman email associated with Lost&Found, all emails are received securely via Google hosted SMTP server Intelligent suggestion system for entries from any user →Using KNN to compare to opposite-flagged items (lost vs. found)

Challenges we ran into

Our team consisted of eight members who specialized in a variety of roles, including front-end, back-end, and product design. Communicating with a large group meant we were able to gather a diversity of perspectives.
Learning Python and an unfamiliar framework
Maps API
Putting together independent features

Accomplishments that we're proud of

Working with a dynamically typed language (most of us are used to writing in C, Java or C++)
Putting together a usable, beautiful front-end

What we learned

Discussing and agreeing on a plan is just as important as coding

What's next for Lost And Found

An additional feature we'd include is a way for users to converse with each other about a lost item through an authentication system or anonymous messaging system. We'd also like to connect the keyword extraction and automatic keyword matching that we implemented independently!

Video Link

https://www.youtube.com/watch?v=VQNu8CyeROs&feature=youtu.be

Built With

Created by

I implemented the Google Cloud Vision API to extract object labels from user-uploaded images. The labels auto-populated a database entry about lost/found local objects.

Coby Simler
I worked on developing the secure email transfer protocol, implementing foundation for anonymous emails between users connected by mutual interest in a lost (and found) item. I laid framework for suggestion algorithm using text-based and numerical K-Nearest Neighbors to suggest a queue of similar entries flagged as the opposite kind of post (i.e. suggests only similar "found" items to a user who has just listed a "lost" item).

Matthew Román
I worked on processing the POST requests from the submission form for lost or found items. I also created the front-end for the website, which includes the index page that has the map with pins as well as well as the form submission page. I used the Google Maps API as well to customize the added pins on the page's map.

Cassandra Marcussen
I helped integrate the Google Maps API into the front-end using HTML/CSS and assisted with sending/retrieving form requests for image uploading via Flask. I engaged in overall product management by coordinating across the front-end and back-end teams to merge the final app.

Annie Sui
I worked on implementing post and image submission, from front-end HTML forms to saving the images on the server.

Lucie le Blanc
I worked on developing the Python Flask application that interfaces with the front-end in managing a database of information.

James Yang
I worked on creating, filling, and reading the database using Python and sqlalchemy. I also designed the logo.

mab2400
I worked on extracting keywords from user's description on the item and finding potential matches from the database.

Bennington Li
Cassandra Marcussen