Sustainability of our dwindling food resources is one of the most urgent concerns facing our ever-growing and developing population. As our consumption increases exponentially, it becomes increasingly important to reduce waste. FoodSaver provides a fast, convenient way to transform leftover food into delicious meals.
What it does
The android app targets the issue of food waste by providing an intuitive interface which recommends recipes to users based on their leftover ingredients at home.
How we built it
We built the app in Android Studio using Java. We implemented computer vision to classify photos of ingredients by creating and training a model using AutoML Vision and Google Firebase. A string of ingredients is returned which is queried to Spoonacular API via Retrofit to find an optimal recipe based on these ingredients.
Challenges we ran into
Some challenges we faced included learning to use Android Studio, creating an accurate yet low-latency and small-footprint client-side model, and interactions between the app and the API.
Accomplishments that we're proud of
- Being able to implement fully-functional object recognition to the app
- Creating a functional app that does what its supposed to
- Creating a clean UI from little/no mobile development experience
- Finding ways to get recipes from the ingredients via APIs
What we learned
We learned a lot about the issue of food waste itself, mobile app development and computer vision as a whole.
What's next for FoodSaver
We want to make it much more interactive with many more features, such as relevant cooking videos, sharing your dinners with others automatically through social media, hooks with local grocers, and releasing on the app store. The sky’s the limit!
For our domain.com entry, we used the domain knickknack.tech :)