Eye Health Predictor

Inspiration

Costly and hard to reach eye health checkup

What it does

Health status of human eye. it can classify and predict the condition of given human eye image input into 5 different categories as follows

1. Healthy eye
2. Cataract
3. Glaucoma
4. Chalazion
5. Pterygium

How we built it

Google AutoML Cloud

Challenges we ran into

There are no predefined data sets readily available for us to work on so more than 70% of time is utilized for collection and cleaning of data. After that we had to tune our model performance to get a better output with every iteration

Accomplishments that we're proud of

Over 90% precision and over 80% recall of ML model.

What we learned

Google Cloud Platform, Machine learning, integration, Rest API, Mobile App

What's next for Eye Health Predictor

Diversifying the data, More disease inclusion, Third party API.

Built With

• google-cloud-automl