We are often forced to make career choices at an age when we cant choose between Cricket and Football . What happens if we make the wrong choice? We can either accept it as it is or steer our career to achieve our goals no matter how tough it is !
Saksham Chaurasia, our next pathbreaker, Machine Learning Engineer, applies computer vision techniques in the fields of Surveillance (facial detection/recognition, object tracking, etc) and Identity Management to address challenges in Image Recognition.
Saksham talks to Shyam Krishnamurthy from The Interview Portal about transitioning from Electronics to Artificial Intelligence after discovering the power of machine learning via the Google I/O developer fest after being exposed to Alpha Go, an AI based program that uses deep reinforcement learning to play the Go game
For students, AI is present everywhere, try to understand how these technologies are complementing human abilities and pushing the limits of what is possible.
Saksham, tell us about your background?
I was born and raised in Gorakhpur, U.P. I come from a middle class family and both my parents are involved in the government. service. I finished high school from a co-ed missionary school in my hometown. Occasionally, I played cricket and football at school competitions.
Growing up, I had a keen interest in science, technology and music which later encouraged me to take up engineering as a profession.
What did you do for graduation/post graduation?
I pursued B.tech in Electronics and Communication engineering from Galgotias College of Engineering & Technology which is affiliated to AKTU.
What made you choose such an offbeat, unconventional and rare career?
I have been a technologist long before I even joined college. During the early days of college, I started tinkering with various technologies ranging from both software to hardware. The influential point in my academic career was when I discovered the power of machine learning via Google I/O developer fest. Through this fest I was exposed to Apha Go, an AI based program that uses deep reinforcement learning to play the Go game and subsequently defeated Go world champions in different global arenas and arguably became the greatest Go player of all time by tackling more than a billion combinations of decision making. This experience got me hooked to AI and there was no turning back when I saw how the power of AI could take on seemingly impossible cognitive tasks.
How did you plan the steps to get into the career you wanted? Or how did you make a transition to a new career? Tell us about your career path
My undergrad degree was a major in hardware but I had decided that I wanted to work in the software industry building intelligent systems. I knew that this transition wasn’t going to be easy. So, I started looking for projects that would make me stand out in the market. I also started to maintain my LinkedIn profile and started connecting with people from the industry.
This was something that I started planning in the 2nd year of my engineering degree. Since AI was completely different to what I was studying I wasn’t sure of how to proceed. So I started reaching out to people on Linkedin to get some advice on how to build a career in AI. The response wasn’t very encouraging.
I decided to give it a shot and started learning the basics in my 2nd, 3rd year. I started by learning Python from UDEMY and then took up a course “Introduction to Machine Learning by Andrew Ng” after doing some research on Coursera. I got really good at it and started looking for some real-world projects. BTW many of these online courses offer scholarships for college students. So they are pretty much free for students. I also took several courses from different MOOCs platforms such as Coursera, EDx, Udemy, etc and as I moved forward my with real-world projects I got fascinated by Self-Driving Cars and other Applications of AI in Healthcare soo I did many courses and projects on them. According to me, learning is a continuous process and therefore I still get enrolled in courses that I find interesting and currently I am taking a course in Nanodegree from Udacity.
At this time I also realized the importance of creating a profile on Linkedin to showcase my work and capabilities as well as taking advantage of a platform like GitHub for projects.
I started by making my Linkedin profile first and uploading details of all the practical projects that i did as a part of my online courses and other projects on GitHub. By the end of 3rd year i was getting several requests on Linkedin to work on short-term projects with little or no remuneration. I was ok with it since this was the learning phase of my career. Though i didn’t get paid for my work I got to apply AI in different sectors and address different problems, which helped me build my profile. On the way I met a few industry engineers who further guided me towards building better skills.
That is when i realised that there are websites such as UpWork where i could actually bid for projects and get paid for work. This was a revelation and I started doing projects there as well. Despite my freelancing assignments, several employers were a bit reluctant to hand over projects to a 3rd year student because of lack of experience. But i did manage to get several short term projects with a meagre monthly salary for 3-4 months.That is something I had to deal with.
During the 4th year of our college we had to work on a final year project in order to graduate. And since i had taken up Electronics, the project had to be related to my degree. While I was looking around for a relevant project, my dad who works in the Agricultural sector and knew that i was working on software, came to me with a problem that farmers in his line of work were facing.I stumbled upon an idea while discussing it over with my mentor that involved building a plant disease recogniser. I worked on it as my final year project and it got selected as one of the best projects in the entire year. That project also helped me secure a biomedical imaging related internship internship at a university in China where I had worked on breast cancer segmentation and also published a research paper on it.
One of the big problems for farmers is loss of income due to plant diseases. By the time they notice the symptoms of the disease, taking samples to a lab and identifying the disease, which sometimes takes a week or 2, the loss is irreparable, also leading to a huge loss of income to the farmer. My father wanted to know if i could use AI to identify the disease well in advance so corrective measures could be taken to protect the crops from destruction.
I had never worked on such a problem before. But, with the help of a mentor from CDAC Ashutosh Singh and my friend & colleague Manaswini Bajpai, we decided to attempt a solution. We looked through several images of diseased plants on google, with 800-1500 images per disease, covering around 30-40 diseases. Based on the images we labeled the disease, we had to spend a lot of time ( 2 months) making sure the data was clean and ready for the algorithm to be trained on.
Our next step was to take the digital images of the plants being analyzed and apply image recognition to identify if a plant had a specific disease and what it was. We were actually able to achieve more than 90% accuracy with our algorithm. Though I had a tough time convincing our faculty to accept this as our final year project, they eventually accepted it.
I had completed my final year project well in advance. And that was when I got my biggest breaks and a turning point. Around the last semester at college, i received an unexpected internship offer from Shantou University – China to work as a AI Researcher Intern under renowned professors for developing a system that could detect and segment a breast cancer tumor. I worked on Breast cancer detection on ultrasound images using image processing and deep learning. We developed the system based on Modified U-Net on Breast Ultrasound Images.
I worked in China among a team of around 5-6 AI professionals, several doctors and under the guidance of 2 professors, where we had to build an Image Recognition system that analyzed around 800 breast images per person, capturing high resolution images of the breast. Of these 800 images we had to isolate around 200 images which contained a tumour and highlight the tumor as Red if it was cancerous or green if it was benign.We also generated a 3D Image of the tumor for doctors to analyze. We also had to tag the precise location of the tumor for doctors to be able to identify. This Deep Learning algorithm was trained on around 200 patients with labels cancerous and non-cancerous tumors.
I published a part of work as a research paper, here are some details of it-
“Nipple Segmentation and Localization Using Modified U-Net on Breast Ultrasound Images”
Breast cancer causes massive deaths every year. To prevent this, the detection of a tumor in its initial stage is necessary. And correctly identifying a tumor from an ultrasound image requires years of experience and due to a few of a number of experienced specialists makes it a challenging task. In this work, a modified U-Net model named GRA U-Net is proposed to assist specialists in acceptably ascertaining a tumor in an ultrasound image. GRA U-Net is a combination of some of the existing techniques and can segment the nipple from Automated Whole Breast Ultrasound (AWBUS) images. Nipple segmentation is important as it can help in precisely locating the tumor from the outside of the breast. The segmented nipple can thus be used to locate tumor with respect to its position.
These achievements made me stand out in the market and when I got back from China I got a few offers. I finally decided to join Signy Advanced Technologies where I’m working currently.
Signy Advanced Technologies basically help consumers leverage the power of AI and Blockchain. We have products that automate the process of Identity management, employee attendance management, and other tasks involving healthcare applications using the secure processing of blockchain.
Other than that, I am working with an Edutech team to guide students who want to get into this field of AI.
How did you get your first break?
I got my first job through LinkedIn. On returning from China, I got offers both on-campus and off-campus. I finally went with the offer from Signy Advanced Technologies where I work currently as a Machine Learning Engineer,
What were the challenges? How did you address them?
Challenge 1: My undergrad degree was a major in Electronics, but I had decided that I wanted to work in the software industry building intelligent systems. I knew that this transition wasn’t going to be easy. So, I started looking for projects that would make me stand out in the market.
Challenge 2: I had no proper guidance during the initial phase of my career.Things were not very clear due to a lack of direction. I had faced a lot of hurdles and used to get stuck for hours or even days!! But C’mon that made me what I am today “A Warrior” 😀
Where do you work now? Tell us about your work
I work at Signy Advanced Technologies (https://signy.io/). Here we have an impact-driven factor which is “Fighting inequality. Creating opportunity. Driving stability.”
“More than 1 billion people across the world have no identity, no access to civic amenities and basic services. We are building identity solutions to help drive the world towards a decentralized system that builds a sovereign, equal-opportunity future for everyone.” using the power of cutting edge technologies such as Blockchain & Artificial Intelligence.
For more details go to signy website.
The skills that are required are proficiency in python and machine learning concepts as well as some of the software engineering fundamentals.
I love the impact that we’re having on the society and that we’re the pioneers in building the homegrown AI and blockchain applications in India.
Currently, we are working on 3 products –
- Signy – Smart Onboarding, Sanction Checks & Identity Verification.
- AttendX – Smart Attendance, Real-Time Monitoring And Activity Log
- Virality – Vitality provides an ecosystem for contact detection, patient information, health data, and data modeling in a safe and reliable manner while maintaining data security.
We primarily use computer vision techniques and Deep Learning algorithms to scan, identify and authenticate IDs such as passports, aadhaar, PAN, etc with very high accuracy. By comparing the images from the Ids with selfies of the ID holders, we confirm and authenticate people. The system is tamperproof in the sense that it is not possible to use videos instead of selfies because our algorithm can detect it. This helps perform authentication in seconds and then we use the power of blockchain which is decentralized systems to store the information of the user(details & IDs) which can be used for onboarding, KYC, checks & identity verification.
Another application of my work in Deep Learning is employee authentication at the workplace based on facial recognition instead of biometrics. Again, based on the ID of the employee, we can perform a match with the person entering the premises, thus validating ID in real time with a touchless interface. This system even works for visitors by validating their IDs through facial match, which again helps in authentication.
How does your work benefit society?
Inequality in access to resources leads to inequality in opportunity, and subsequently inequality in outcomes. This vicious cycle affects lives and careers across the globe and also harms the environment. Signy is committed to solving these issues and harnessing the power of emerging technologies towards Sustainable Development Goals (SDGs).
According to our Product Manager Mr.Ashish Kots – “More than 1 billion people across the world have no identity, no access to civic amenities and basic services. We are building identity solutions to help drive the world towards a decentralized system that builds a sovereign, equal-opportunity future for everyone.”
Talking about AI – According to Andrew Ng (Well he is one of the pioneers of AI) – “AI is the new Electricity”, Just as electricity revolutionized lives 100 years ago, AI is doing the same right now!! AI is present everywhere you look around – from our phones(alexa, siri, etc) to healthcare, autonomous cars, etc and it is just the tip of the iceberg, we have not even fully explored its applications yet!!
Tell us an example of a specific memorable work you did that is very close to you!
My most memorable moment is the time when I was working as an AI research intern at Shantou University China. It is memorable not only because of work but various other reasons also like-
- It was my first major breakthrough in my career.
- It was my first international trip and gained a lot of experience working with a team which consist of persons from different nationalities
- Biomedical and self-driving is my area of interest and since the whole project was based on biomedical, I totally loved the idea and working on the project.
- I learned many important concepts/lessons on AI, healthcare, a bit Chinese language & their culture.
- Most importantly, I made some good friends, had much fun, went hiking, swimming, learned cooking, etc!
Your advice to students based on your experience?
- I would advise the students to explore their fields of interest and always keep in mind that you need to work both hard and smart to get from where you are to where you want to.
- Sometimes things don’t work out as planned and you don’t know what to do and how to proceed, but believe me there is always a way, just don’t let failures demotivate you!
- Never stop learning, learning is a continuous process, the day you stop learning you’re doomed.
- Money is important but don’t let money overpower your passion.
I don’t have a concrete 5-year plan right now but I do have some ideas which I would be doing in next 5 years-
- I wanted to pursue research in the field of applied machine learning but due to a sudden pandemic outbreak I think I’ll skip that part and for now I’ll focus on improving my skills.
- I love working on AI biomedical and healthcare and have some interesting ideas soo most probably I will be working in that direction in near future.
- Cover all most all the parts of India!!! I love traveling!