Inspiration
- For the health fanatics, fitness enthusiast or foodies it helps to know what you're eating
- Most of us not eating healthily in lockdown (be honest, you aren't)
- Want to have a new way of tempting ourselves to eat healthier
- Also learn more about nutrition at the same time
What it does
- Uses your mobile phone camera to detect foods using computer vision and displays 3D models of nutritional value in augmented reality
- Easy to use. Just point your camera at a fruit and try!
- Family friendly. Children can use it to learn. Suitable for all ages.
How we built it
- Google Cloud AutoML Computer Vision for recognising fruits
- C# and Unity to display 3D models stored with echoAR
- MATLAB to build 3D models from nutritional information
- (Additional) Java and Android Studio for Android version
Challenges we ran into
- Little prior experience working in Unity / Android Studio
- First time of using echoAR
Accomplishments that we're proud of
- Training a ML model to distinguish different types of fruit
- Learning to incorporate multiple APIs and cutting edge technologies
- Displaying models in augmented reality space using Unity and C#
- We had fun building an interactive fully functioning app (some for the first time)
What we learned
- Choosing the right platform
- Quickly picking up new frameworks/libraries and putting them into use
- Learning to use ML models and google cloud computer vision API
What's next for Fiberr
- More types of food
- Alternative designs
- Faster detection rates
Log in or sign up for Devpost to join the conversation.