Anonymous

Inspiration

From the start, we were very interested in learning and applying ML/AI concepts to real-world problems. We want to explore the world of AI/ML and are very enthused to learn and grow further. This hackathon has given us a wonderful opportunity to do the same :)

What it does

Our project is an image detection software that detects and evaluates handwritten mathematical equations, to give an integer output, based on the given expression.

How we built it

We had built our model using PyTorch. It was the first time we had ever used PyTorch(or any other ML software for that matter), so we spent some time looking at how the syntax works and did our best to make an ML model that fitted the necessary requirements.

Challenges we ran into

Since it was our first time coding an ML model, we struggled quite a lot at the beginning, as we were getting to grips with the syntax. Also it was quite tough initially to come up with a good plan to try and tackle the problem statement. Nevertheless, I feel we did our best, and would hopefully do better as we go deeper and deeper into the world of ML.

Accomplishments that we're proud of

We feel that the hackathon enabled us to come a long way in a short span of time, and we got a good hands-on experience that would stand in good stead, while building other models. I would like to thank AIMLC IIT Delhi, for providing us with a wonderful opportunity to learn and apply ML concepts.

What we learned

It was our first time doing an ML project. The problem statement was quite interesting and challenging. We learnt a great deal about the primary ML libraries- PyTorch and TensorFlow/Keras and how to implement them in our models to solve the problem statement effectively. More importantly, it also helped us understand where we might have been lacking, and how we should overcome our shortcomings to learn further. We hope to understand and learn further in our ML journey, and apply concepts to help this model(and other models we may make in the future) have a better accuracy and more robust :)