The domain of Machine Learning has found applicability in almost all spheres of life. From smart voice assistants to detecting COVID, Machine Learning is everywhere. In this hackathon, the aim is to apply ML to develop an application in the field of Optical Character Recognition. Given an image of a math expression containing a symbol (among +, -, ×, ÷) and two digits (among 0-9), you have to create a ML model that can predict the result of the math expression. The math expression may be in the format of pre-order(symbol before digits), post-order(symbol after digits) or in-order(symbol between digits). The problem is divided into two tasks:
For an image:
- Predict the expression style (Label): In-order, Pre-order or Post-order [30 pts]
- Predict the result of the expression (Result) [70 pts]
Bonus: There are 10 bonus points for each task for not using any external data or pre-trained models to solve this problem.
For more details, read this information brochure
Download the dataset required for the hackathon from google drive. To know more about the nature of the data and annotations, refer to the dataset