As students, we are faced with the challenge of preparing meals and buying groceries, on a tight budget. We envisioned a tool that would provide a smoother grocery shopping experience.
What it does
Cart-astic gives you the ability to efficiently plan delicious meals by recognizing food items in images you take and helpfully suggests recipes based on your virtual cart. It keeps track of which ingredients you still need to grab, and which dishes you have enjoyed in the past to make your life even easier.
How we built it
- We used Microsoft's Cognitive Services API and image recognition to determine a food object from the user's camera snapshot.
- We then built a comprehensive and intelligent microservice (with features for optimizations, such as request caching) to return a wide range of recipes that include information such as nutrition, price breakdown, etc. that the user can dynamically filter.
Challenges we ran into
Our team found that the ability to recognize food items using computer vision was still in its early stages, compared to technology like facial recognition, which meant that we could not identify the full range of foods.
Accomplishments that we're proud of
Building a superior user interface with a very simple and fluid workflow with intuitive design. This resulted in an enhanced user experience, which allows users to return to the application to continue making use of its benefits.
What's next for Cart-astic
Using data to further enhance the user experience by providing targeted coupons, better recipe suggestions based on other users' preferences, and creating weekly meal plans with budget restrictions. This will empower the community by optimizing resources and saving the users' time and money. It would also benefit grocery stores as it will eliminate waste by sending targeted deals for soon-to-expire food items.