6 Computer Vision Research Skills UC Berkeley Electrical Engineering Majors Develop: Ethan’s Story

Computer vision algorithm research uses machine learning and neural networks to teach computers how to meaningfully process and analyze visual data so they can engage in various decision-making processes. This field has a lot of potential and is actively being developed in order to advance automated cars, facial recognition, surveillance, medical imaging, robotics, and more. By inputting large sets of visual data into various computer programs and software, electrical engineering and computer science majors are able to train computers to “see” images and scenes more clearly. 

Ethan Kou, a Polygence alum and first-year student at UC Berkeley, is studying Electrical Engineering and Computer Science (EECS) on account of his passion for robotics, machine learning, and computer vision. During his time with Polygence, Ethan was able to gain a strong foundation in computer vision, which has allowed him to hit the ground running in terms of his undergraduate research. In fact, Ethan is currently a Berkeley Artificial Intelligence Undergraduate Researcher, as well as a Computer Vision and Deep Learning Intern at Matterport. He also won first place in Machine Learning at the Berkeley hackathon — and he’s done all of this in his freshman year!

The UC Berkeley electrical engineering and computer science program is ranked among the top three undergraduate computer engineering programs in the world, which further distinguishes Ethan’s accomplishments. It also leads us to an important question: How can you start developing the essential computer vision research skills needed for UC Berkeley?

In this article, we’ll discuss 6 computer vision research skills that UC Berkeley’s electrical engineering majors develop. We’ll also discuss how high school research programs, such as Polygence, can help you stand out from your peers during the college admissions process, giving you a shot at being admitted to UC Berkeley — just like Ethan. 

Overview of Core Computer Vision Algorithms

Computer vision relies on a wide array of algorithms designed to enable computers to process and analyze images and digital data. These algorithms form the backbone of tasks like image recognition, object detection, and automated scene interpretation. At a high level, computer vision algorithms can be based on traditional techniques, such as edge detection or feature extraction, or on modern deep learning architectures that leverage neural networks for more complex tasks. Understanding these algorithms is crucial for students in EECS courses who want to design and implement real-world computer vision applications.

Why Computer Vision Research Is Key for EECS Majors at UC Berkeley

Combining electrical engineering with computer science is a great way to position yourself at the forefront of cutting-edge technologies, such as autonomous vehicles, AI, and robotics. Not only will this college major increase your job marketability, but it will also prepare you for completing meaningful and in-depth research throughout the course of your career. 

Ethan Kou discovered the importance and marketability of combining these two disciplines upon completing his first of two Polygence projects, State Estimation and Motion Planning for an Autonomous Competition Robot. Ethan shares that “My project focused on improving perception robustness in bad weather for self-driving cars. In bad weather, the camera quality gets worse, which affects perception. I wanted to make perception more reliable in those conditions.” 

Ethan continued to reflect on his Polygence experience: “That project really sparked my interest in machine learning and computer vision. From that experience, I learned about different computer vision algorithms that I found interesting, and I’ve been able to explore them more at UC Berkeley.”

While computer vision algorithms and applications may not be a part of a standard high school curriculum, Ethan leveraged his time with Polygence to get the necessary background, technical skills, and confidence to explore this discipline before even stepping foot on UC Berkeley’s campus. After all, computer vision research is key for EECS students at UC Berkeley, so getting a head start is a great investment in your future!

Common Challenges EECS Students Face in Computer Vision Research

It shouldn’t come as a surprise that computer vision algorithm research is full of challenges. In fact, Ethan believes that accepting failure as a key part of the learning process has been crucial to his success during his first year at UC Berkeley. 

In addition to learning how to manage your time effectively by balancing coursework, research, internships, and extracurricular activities, EECS students are tasked with keeping up with rapidly-evolving research and technologies. EECS students must also be able to understand complex algorithms and mathematical models. 

From collecting large quantities of quality data and mitigating potential biases to knowing how to set up and troubleshoot complex hardware and software systems, EECS majors certainly have their work cut out for them!

Ethan’s 6 Essential Computer Vision Research Skills

Ethan’s UC Berkeley electrical engineering education is strengthened by his computer vision research skills — skills he began forming with Polygence and his two research program mentors while he was still in high school. Computer vision algorithm research is complex and therefore requires a strong foundation, persistence, technical skills, and the ability to engage in creative problem-solving. In this section, we’ll review the 6 essential computer vision research skills needed to succeed in EECS, as well as strategies to cultivate each of these 6 skills. 

1. Fundamentals of the Research Process

Strong computer vision research begins with a solid understanding of the research process. This includes defining a clear problem statement, reviewing existing literature, formulating hypotheses, and designing experiments. Working with a research mentor is often the best way to gain these skills, especially if you’re in high school.

Ethan’s introduction to research through Polygence was inspirational. In fact, he remarks that, “After my first project, I gained a lot of knowledge and even published a research paper. That experience motivated me to pursue a second project to build on what I had learned.”

2. Algorithm Development and Optimization

Writing computer algorithms means knowing how to turn your ideas into working code, choosing the right methods, and making sure your solution is accurate and unbiased. Optimizing algorithms goes beyond basic coding, too. After all, even the smartest algorithm isn’t very useful if it’s slow or uses too much computer power. That’s why learning how to tweak and troubleshoot your code so that it runs faster and works well on different types of hardware is just as important as writing the initial algorithm. 

3. Technical Reading and Writing

Interpreting a research paper goes beyond just understanding the abstracts; you’ll need to be able to break down methodologies, understand mathematical formulations, and assess experimental results for validity. On the flipside, you’ll also need to learn how to document your own research findings in clear, precise language. Ethan shares that, “The technical skills I learned [at Polygence] have been useful in my classes and research here. Reading and writing papers is something I do a lot in my research lab, and that background definitely helped me.” Overall, knowing how to read and write scientific papers is the cornerstone of a successful STEM career.

4. Machine Learning Techniques

Computer vision relies heavily on machine learning to improve perception and recognition. This means you’ll have to learn the basics of machine learning, like how to train a model, what loss functions are, and how different types of neural networks work. For more information on machine learning, check out our article, Machine Learning for High Schoolers: A Comprehensive Guide. 

5. Creative Problem-Solving

Ethan’s Polygence research mentors helped demonstrate to Ethan how his early interest in art informed and ultimately inspired his interest in STEM. After all, creativity plays a crucial role in STEM research. Ethan remarks that “Polygence allowed me to think creatively about different ways to approach a problem.” Creative problem-solving is a valuable skill because thinking outside the box is the best, and often the only, way to meaningfully address research challenges. 

6. Persistence and Ambition

Staying motivated through setbacks and seeking new opportunities leads to personal and professional resiliency. After all, nothing worthwhile comes easily. You have to work hard, make sacrifices, and be patient with yourself as you approach new challenges and familiarize yourself with new material. 

7. Real-World Applications of Computer Vision

Computer vision is not confined to academic exercises; its applications span numerous industries. Detection and recognition algorithms power autonomous vehicles by identifying pedestrians, traffic signs, and obstacles. In healthcare, image-based algorithms assist in analyzing medical scans for early disease detection. Robotics relies on deep learning and digital vision systems to navigate and manipulate objects. By connecting algorithmic concepts to practical applications, students gain a better understanding of how theory translates into impact.

How Ethan Navigates UC Berkeley’s EECS Environment

Ethan’s ambition has only continued to grow now that he’s at UC Berkeley. He’s always been a motivated student, but in college, he remarks that “the classes cover topics I wanted to learn in high school but never had the chance to study.” He’s also keen to take advantage of interdisciplinary classes in electrical engineering and computer science, reflecting that, “I’ve always been interested in robotics, and EECS is a great major for combining the hardware side of electrical engineering with computer science.” 

Outside of the classroom, Ethan is just as busy, engaging with clubs and research labs focused on robotics and AI. And the best part about his extracurriculars? The fact that “there are also many clubs where you can meet driven people who share the same interests.” For Ethan, UC Berkeley is the perfect place to engage in meaningful research and seek out mentorship and camaraderie with other EECS majors. 

Advice for Prospective EECS and Computer Vision Students

Ethan has a lot of advice for prospective EECS students in regards to getting accepted into UC Berkeley and continuing to succeed after enrolling. 

It’s no secret that getting accepted into UC Berkeley is difficult, especially as an EECS major. For Ethan, the answer is simple: “The most important thing is to have one thing you’re really passionate about and keep pursuing it. For me, that was research projects like Polygence.” He continues by saying that “If you’re in high school and really interested in a topic that isn’t taught at your school or you’re not sure how to explore it, research is a great way to dive deep into that subject.”

Ethan also believes that perseverance and accepting failure as a key part of the learning process have been crucial during his time at UC Berkeley: “Once you’re at Berkeley, it’s important to be hard-working and to keep going, even when things are tough or you face setbacks. It’s also important to be ambitious and look for opportunities because at Berkeley, there are many, but you have to seek them out — they won’t just come to you.”

Preparing for a Future in Computer Vision and Robotics

For better or for worse, students interested in machine learning and robotics programs for high school students need to pursue their interests largely outside of the classroom. After all, standard high school curriculums don't usually include a lot of machine learning, data science, engineering, and robotics. This is why it’s important to seek out AI internships, data science research opportunities, engineering competitions, and other opportunities for high school students. 

Polygence offers a challenging and comprehensive Research Mentorship Program for motivated students, like Ethan Kou. There are lots of benefits of having a mentor in high school and Polygence prides itself with pairing students with their ideal research mentor — someone who shares common interests, learning style preferences, and professional goals. 

Not only will developing research skills in high school strengthen your future academic prospects, but it could also open the doors to various industry roles after college. So whatever your source of motivation may be, developing research skills while you’re still in high school is a great investment in your future.