SLIDES
https://docs.google.com/presentation/d/1gNpM895Vu57GfwlDTGxhGcQtKBnTYCEsFzj1mhLuwRk/edit?usp=sharing
Inspiration
While NYC is filled with opportunities, we all are aware of homeless problem. We want to help them back on their feet. One way to do it is to give them all they need at least. Lot of us want to help but aren't always at right place and position.
What it does
It lets people take picture of homeless person and app finds them in database using facial recognition API from IBM Watson, then it shows list of things they've requested. You can also add new things you think they need or ask them what are they in need of. App uses GPS as well to show homeless shelters nearby and their phone numbers, and also show what other homeless people are in this area.
You can also search by name.
How we built it
- nodejs
- javascript
- watson api
- open data api
- html/css
- git
- heroku
Challenges we ran into
- Training AI model requires good amount of data, our initial plan to upload a picture and have it search in database needs to revisited.
- Lack of time, idea took lot of refining.
- Sleep :D
- learning open data api
Accomplishments that we're proud of
- we finished in time
- we came together as group
- we integrated with 2 external apis (watson, open data) AND with html web apis ( webcam and geo location)
What we learned
- we learned things about AI based visual
- we learned to appreciate open data
What's next for HomelessTracker
- release version 1.0
- work on launching with non profits
- fix code
Built With
- ibm-watson
- javascript
- node.js
- open-data
Log in or sign up for Devpost to join the conversation.