-
basic idea sample
Inspiration
'SuperQuest' means, a quest in supermarket, looking for some nutrition.
More Children are getting obese and we feel obliged to educate them with nutritional facts of food they eat every day while implementing it in a fun way.
If they get very basic knowledge in gamified app, such knowledge can contribute to healthy food life in adulthood.
What it does
Basically this app is a lind of general object recognition, especially for food. This can overlay nutrition information(or just coloring is OK for little children), or country which the food come from.
1) Utilizes Spponacular API, https://market.mashape.com/spoonacular/recipe-food-nutrition#quick-answer Due to the json answer is in a full english sentence, we used regex and parse the strings, which return only the nutritional value. *Special thanks to the head up from Spoonacular.
2) Participated in HelloFresh Challenge - Data nutrition value from USDA National Nutrient Database (Strawberry and grape, which is not available in Spoonacular API). Also we implemented the ingredients.csv file, by utilizing the country data, which showed the country origin of the fruits.
How we built it
We separated our task, Toshi focused on deep learning and fruit image recognition, while hunter built the front end and fetch the data from API.
Challenges we ran into
Time Accuracy in food object recognition.
Accomplishments that we're proud of
Fed the learning machine and built deep learning system by ourselves
What we learned
Work harder and smarter!
What's next for Super Quest
Built cross platform mobile Apps (Android and iOS) and then turn our dreams into real world!
Built With
- chainer
- deep-learning
- image-recognition
- machine-learning
- python
- webrtc
Log in or sign up for Devpost to join the conversation.