As students in STEM, we have had more than enough experience trying to solve matrices on Symbolab or on other calculators. We felt that the process of inputting matrices could be streamlined using image recognition via a neural network. We designed an app that --upon scanning an image of a matrix-- would turn it into reduced row echelon form. The app was written in flutter with back-end calculations done in python. Since most of the team lacked experience working with neural networks and flutter we had to spend a good deal of time researching possible solutions and methods of implementation. Currently, the app can scan 3-by-3 matrices of single-digit numbers and solve them. In the future, we would like to implement more variety in matrix size, allow for multi-digit numbers, and more complex matrix operations.
Matrix Scan
Inputting matrices into calculators or online tools can be a difficult and time consuming task. Our project simplifies the process by using neural networks to recognize numbers and solve matrices.
Updates
Leave feedback in the comments!
Log in or sign up for Devpost to join the conversation.