Located adjacent to CMKL University, King Mongkut’s International Demonstration School (KMIDS) is Thailand’s first international demonstration STEM high school. KMIDS produces high quality students for KMITL and other leading universities in Thailand and abroad. With programs influenced by some of the top universities in the world, such as the ‘STEM with Robotics’ program from Carnegie Mellon Robotics Academy in the U.S., KMIDS is a leader in innovation. Under an internship program with CMKL, three KMIDS students assisted in ongoing research projects regarding Machine Learning. The students worked with researchers over the course of three months; each student choosing a specific topic and conducting research. Machine Learning is a form of Artificial Intelligence (AI) in which the program is designed to learn on its own. Through the study of algorithms and statistical models, patterns are detected. This information is then applied to computer systems and used to perform specific tasks without explicit instructions.
One KMIDS student, Bank, decided to innovate a method to increase the security at his school. Due to the lack of security precautions, he searched for a simple way to create a safer environment, while still allowing ease of access for students and faculty. Bank decided on facial recognition, a process that is conducted instantaneously without causing a delay. The objective of this system is to identify people entering the school’s gates, as well as search for specific people within the school. Bank’s system is currently a work in progress; however, he is optimistic that this can be an effective security system that can be implemented on a larger scale.
Interviewer: What are your thoughts on machine learning?
Bank: At first, I thought it was an interesting topic, but I was still skeptical about it. I didn’t completely understand it prior to working with CMKL, but I now feel that I have a base knowledge regarding the topic. Once I got a grasp of how machine learning can be applied, I see how useful it can be in our everyday lives. I think it can be revolutionary in the near future, maybe within a few years.
Interested in smart cities, Ping wanted to create a program that was able to recognize buildings from a satellite-view image. Using a Python package designed to collect information of buildings from OpenStreetMap, as well as a package designed to collect satellite image tiles from Bing Maps, Ping gathered training data for a neural network. He then used Mask R-CNN, a type of neural network designed to identify different objects in an image, to train the data that was gathered.
Interviewer: What specific skills did you develop or improve upon from working with experienced professionals at CMKL?
Ping: I believe my analytical skills made the greatest improvement. Programming requires a lot of critical thinking and alternative solutions in order to solve complex problems and I am now much more comfortable trying different approaches. I used programs such as Python which required great attention to deal. While other programs use notations, Python uses indentation, so you have to be careful when it comes to spacing.
Jeew opted to design a system utilizing object recognition, focusing specifically on smart farming. Using picture to training data for a neural network, he was able to train a computer system to determine whether animals were pigs or piglets. Although much progress is needed until the project is complete, Jeew is optimistic that this is a technology worth pursuing further. With the ability to adapt to a variety of fields from security to farming, he believes object recognition is one of the most applicable outputs of machine learning in our modern world.
Interviewer: Has this experience given you a better understanding of what career you wish to pursue or maybe which careers you’ll try to avoid?
Jeew: For me, I think I’m still searching. This is just one area I can choose to pursue, and I think further exploration is needed. It’s a broad field that is rapidly expanding and this opportunity has given me an introduction to how research projects are actually conducted. Although now that I understand the potential applications of machine learning, it’s definitely something I will consider in the future.
This collaboration between KMIDS students and CMKL researchers provided invaluable experience for the students involved. While overseen by senior researchers, KMIDS students were given a substantial degree of freedom regarding their topic of study. Tutorial instructions were given, and the students were allowed to engage in a form of self-learning that CMKL believes is crucial to the development of analytical and critical thinking skills. As CMKL prides itself on providing experiential, hands-on learning, the students gained unique perspectives as both independent researchers and research assistants. This early introduction to physical learning in the lab will provide insights to how research projects are conducted and also help them determine which particular discipline they wish to pursue in higher education. This collaboration between CMKL and KMIDS was a success for all parties involved. While CMKL benefited from engaged students who made valuable contribution, KMIDS students were given a unique introduction to the world of research at an early stage in their academic careers. CMKL looks forward to future endeavors with KMIDS students and hopes to expand the program as the university expands and grows.
