AR Scavenger Hunt

Target Item
Found Item
Scan Item
Game Over

Inspiration

Scavenger Hunt is fun. But currently, you need to make an appointment with your friends for the time and place, which is not convenient. And now powered by machine learning and augmented reality, we want to try to change it.

What it does

This AR Scavenger Hunt lens allows you play scavenger hunt with friends and compete anytime anywhere. It has competitive gameplay. And you can share with friends your scores and videos of funny moments.

This lens requires you to find as many items as possible within 20 seconds. Each time it randomly lists one item. You need to find it and click the button to scan it for confirmation. Then it'll list the next item. After the game time is up, it will list the total number of items you found, and the names of the first three items.

How we built it

We trained our own machine learning model of multi class classification. Then built competitive gameplay and smooth user experience based on it.

Challenges we ran into

We ran into some challenges when training the models and hooking up models with gameplay logic in Lens Studio.

Accomplishments that we're proud of

It's our first time to create a lens based on machine learning. We're happy with the progress we've made so far. And the final result of our lens is also very good.

What we learned

Now we have a deeper understanding of the machine learning model training process and the lens development process based on it. At the same time, we also see the huge potential of machine learning-based lens in the future.

What's next for AR Scavenger Hunt - Lens Roadmap

Phase-1: To increase the user base and the number of user video sharing, we plan to improve gameplay and add more object classes to the machine learning model. Phase-2: To work with brands, we could add branded products to the machine learning model and ask the player to find brand products in their current environment. Phase-3: For monetization, we would like to see if possible to charge a small fee when users create and share videos based on our lens.