Inspiration
Studying logic and lacking a good way to verify that our hand written proofs are correct.
What it does
It allows the user to scan a handwritten natural deduction proof (or import an image already taken) and the app recognises the characters and symbols. These are parsed into a string which is then parsed and verified in further steps to confirm that the proof is correct or to inform the user which line contains a mistake.
How we built it
Using Android Studio with kotlin for the front-end and UI/UX. A model trained using tensor-flow in python which was then imported into our app.
Challenges we ran into
Training the neural network to recognise with good accuracy
Accomplishments that we're proud of
Making a neural network to recognise characters and symbols. Learning Kotlin and Android app development. Building a strong team
What we learned
How to verify a proof using code Solving problems against the time
What's next for NaturalDeductionScanner
Predicate Logic, shifting the workload from the server to the application, improving some of the UI features.
Log in or sign up for Devpost to join the conversation.