Michigan Hand Map

Finding where someone is from can be confusing. However using the Michigan Hand Map, we put name on the location of where someone may say they are from, just based on where they point on their hand!

Comment

Inspiration

As a Freshman at Michigan State, I found when meeting new people that us native Michiganders love using our hand to show where we are from. However, this can get awfully awkward for our out of state students, as they aren't as aware of the layout of the Michigan cities.

Solution

The Michigan Hand Map perfectly solves this problem. Users can easily input the location of their home based on their hand gestures, and our website will return a name to that location!

How it was built

I trained a Google Teachable Machine with 10,000+ images of users pointing at locations on their hand, and assigned value names to those locations. Then I imported that into my HTML page and used CSS to make it look pretty :)

Challenges ran into

Working on a Hackathon project solo brought me different problems than I expected. I found that I ended up spending much more time polishing up the project and putting in the fine details than I did building the heart of the program. I also chose a project way outside my comfort zone, as this project had much more complex JavaScript than I have ever used, and also this was my first time using machine learning and an API in my program. When programming the JavaScript, it was complex to show the Teachable Machine video into my HTML document. I had to learn much more than I knew to get this stuff done.

Accomplishments in the process

I went past my comfort zone in JS, and used pre-existing knowledge of CSS to work around/with the JS.

What we learned

I learned so much about Machine Learning, as this was my first experience with it. I also learned more about JavaScript and having a video output onto my HTML page.

What's next for Michigan Hand Map

Michigan Hand Map has a huge potential for Education, as young children are known to love using their hands to learn hands-on. Kids could learn basic knowledge about where they live, where the state capital is, and their favorite places around their state. according to a report by the consulting firm McKinsey & Co. Elementary school students in the United States ended the 2020-21 school year four to five months behind where they normally would have been in academic achievement. In order to catch these kids up we need to create better technologies to aid these children in learning faster.

Also, with more time and understanding of tensor flow we could make finding where the user is pointing much more precise, and finding a way to mesh the hand into the wanted state shape, and find the closest city above a population X to the given point. This could work for all 50 states and even countries.

Built With

Share this project:

Updates