Amit Thakur exploring ideas every single day

I have started learning AI

Yes, I have started learning AI, very slowly, and bit-by-bit.

I did a research project in computer-vision during my final year (2013) of college. It was primarily a combination of OCR and object recognition problem. I solved the object recognition part using heuristic approach but had no clue how to solve the former. I read some research paper on OCR and tried implementing them using OpenCV APIs I but didn’t have any clue what algorithms were used and how they were implemented.

At the time, deep-neural-network usage were documented on the web, and some of the today’s popular tools were not there then (TensorFlow) or were just evolving (scikit-learn).

I finished the OCR implementation using Google’s Tesseract. I had EE background and knew basic programming in C++ and worked on GUI using Qt but I didn’t know the better parts of linear algebra, statistics, and neural network to understand advanced ML.

Few months later I got a job offer from Infosys for a Systems Engineering role. I didn’t work on the computer vision projects or any kind of research projects for next 4 years. During these years I learnt enterprise/e-commerce/cloud software application development.

One weekend evening in early summers of this year, I was browsing through my college projects directory, and thought of going back and solving my OCR problem with my own implementation. The ML/AI ecosystem had evolved much in these years. I thought of catching up with the growing AI world. I immediately googled some good MOOC courses on AI. Udacity’s nanodegree on Foundations of Artificial Intelligence looked promising. I immediately signed-up to get the basic understanding of evolving system. I finished the nanodegree before deadline and graduated in June, 2018. It introduced me to Python language, data structure and algorithms, linear algebra, statistics, basics of data science, and basics of ML and AI.

I wanted to learn more. It thrills me to create something cool on my own. So I googled some high quality books, blogs, and open source projects available on AI/ML. This time I wanted to learn all of the nanodegree’s topics in-depth. So I made my mind to learn these slowly and bit-by-bit. I have started reading and doing hands-on exercises.

Let’s see how this journey goes.

Cheers!