As soon as I sat for my A-Levels, I wanted to shift my focus to programming. I wanted to make my one-and-a-half-year stay at home after A-Levels successful. I aimed to learn many things. AI was a booming field. At that time, there were no ChatGPT and DALL-E-like generative models. What caught my interest was the ability of computers to identify objects in real life. At that time, this seemed like magic. I wanted to learn this magic. That’s when I decided to learn Machine Learning. At that time, I did not know any programming language (other than the Pascal programming I did in Grade 10 at school). So, I started learning Python and JavaScript as my first programming languages. My focus shifted from one to the other over time.
Before my A-Levels, I saw a playlist on Computer Programming 101 by MITOCW. As soon as my A-Levels were over, I followed that playlist and learned Python. Still, I could not do much with Python because the videos were purely theoretical. I found the CS Dojo YouTube channel. During the break period, a couple of YouTubers were my mentors, guiding me for the next few months. I started doing a Machine Learning course on Coursera, but I got stuck in the middle. I did not have anyone to reach out to. There were seniors at the university, but I was too shy to ask them for help. That’s how I stumbled upon this blog (https://ramith.fyi). I read his entire blog site and discovered a free premium mentoring program named ScholarX. (https://sefglobal.org/scholarx/2023/)
At this time, I was in awe of photography. I was really impressed with my smartphone camera and how it simulated a real camera. That’s when I realized that I wanted to dig deep into something related to computer vision, specifically computational photography.
During my first year at university, I applied for the ScholarX program. I was assigned a mentor, though not the one I had applied for. (One of my seniors said that this was normal and fine.) I want to share my unfiltered experience. Luckily or unluckily, that mentor did not reach out to me. I tried reaching out to him but got no response. Then, about one or two months after the program began, I was assigned a new mentor who had recently become a mentor in ScholarX. That’s how I got to know Mr. Haritha Jayasinghe. (https://haritha-j.github.io/) He immediately reached out to us and set up a meeting. During a Friday evening in the third semester, we had our first meeting. There were two of us assigned to him, including a final-year student at UCSC.
In the first week’s meeting, we introduced ourselves and shared what we had been working on. He asked about our passionate fields. My partner in the program also shared her final year project. I learned from her as well. Her project was also related to the computer vision domain. By this time, the Generative AI in Computer Vision domain was booming. During the first two weeks, he explained how I should start learning about computational photography, which is a very niche field in computational imaging yet touches a lot of people in the world.
He emphasized the importance of learning the fundamentals in machine learning and computer vision if I wanted to pursue a career in this domain. At the same time, he stressed the need to stay at the cutting edge of the field.
After several weeks, I had read through different resources and started working on a particular subfield in computational photography: Low Light Image Enhancement. If you take a photo at night, you can see the grainy sections of the image. He asked me to build an autoencoder and see how my results would be. I built a neural network for denoising images to remove graininess. During this time, I learned the fundamentals of digital signal processing in the Signals and Systems module, which was helpful for grasping the basics. My learning approach was both top-down and bottom-up.
Next, I started working on the task of brightening low-light images. While doing this task, I learned a lot, from CNNs and autoencoders to GANs. My mentor's guidance was crucial in achieving the desired results.
Additionally, during these sessions, he shared his PhD experience and other life experiences at Cambridge University.
During the last couple of months, I achieved a considerable amount of new knowledge in the computer vision domain and deep learning.
At the time of writing, I have gained valuable insights from him on research conferences and computer vision research. His advice on the research projects I was involved in has been instrumental, and I believe his guidance played a crucial role in helping me secure a highly relevant research internship.
There was one particular thing he said that stuck in my mind: “There are a lot of people out there who are willing to help you if you just ask them”. This applied to lecturers, friends, seniors, or any other person. He emphasized the importance of well-being, noting that being productive is not the only important thing; there should be a balance between well-being and work. He also stressed the importance of physically attending lectures.
I certainly learned a lot from my mentor, not just academic, research related facts but also life-related lessons. I am truly grateful for my mentor. Allocating time weekly to guide someone else — isn’t that great? That’s why I am truly grateful for my mentor in ScholarX. Peer-to-peer mentoring was invaluable, especially when I didn’t know where I was going in life or academically. It helped with personal development and changed the way I see the world.
Here are some articles that I believe can give you more insight into ScholarX.
