Inspiration

Inspired by our various travels, we wanted to build a travel app that uses machine learning to analyze an image of a favorite activity, and suggest places in your local area to visit based on the same interests, as well as similar places around the globe to visit. We also wanted to make it accessible regardless of language ability, while learning machine learning and android development.

How It Works

If you love record shops, ice cream, soccer, and breweries, you can upload a picture of these activities and find local recommendations to curate during your travels and to inspire future travels. This way, you can find your favorite activities, foods, and interests in the area, and be inspired to travel more.

How We Built It

During the hackathon, we all hoped to learn Android Development and Machine Learning. Thus, we chose the Google Cloud Platform (specifically Places API, Firebase, Cloud Vision, and the Machine Learning Kit) to implement our project and integrate with Android Studio written in Java language. We also wanted to do this project based on images so that you can track your new found interests with photos from your travels, and so that it is accessible to people of all languages and speaking/hearing abilities, with low cultural or ability bias.

Challenges

During the project, we ran into challenges learning how to integrate the full stack, and learning how to implement a new HTTP Search Request, since some of the server request clients had deprecated in the past year. Since none of us had android mobile development experience, nor machine learning experience, we had a steep learning curve in learning these new technologies and implementing them. The hardest part was understanding and building an android application from scratch with no prior experience, and using the APIs within the application. However, we feel inspired to continue android development projects now that we have a base of understanding for android mobile development.

Future Prototypes

In the future, we hope to implement more user capabilities, such as marking favorite places on a map interface, machine learning to guess new places you would like based on previous travels and uploads, and profile personalization.

Built With

Share this project:
×

Updates