Answering your most pressing questions about the newComputer Vision Specializationon Coursera are Radhakrishna Dasari, Computer Science and Engineering Instructor and Dr. Junsong Yuan, Associate Professor of Computer Science and Engineering and Director of the Visual Computing Lab from the University at Buffalo, The State University of New York.
An integral part of this Specialization is the use of MATLAB, a leading mathematical computing software for engineers and scientists. Dr. Brandon Armstrong, Senior Online Content Developer at MathWorks, will be discussing how learners can apply and build skills in computer vision using this industry standard tool.
Instructors Radhakrishna and Dr. Yuan: Vision is one of the key senses through which humans learn and navigate through this world. Human vision and the brain evolved over millions of years, enabling us to perform many visual tasks quite effectively. Over the last few decades, it’s been challenging to make a computer see and analyze the visual world as we do! We’ve seen significant progress, however, in the past few years because of the rapidly developing interdisciplinary field of computer vision that adds intelligence to imaging. Its main goal is to automatically understand and interpret images and image sequences. Computer vision has numerous practical applications that are starting to permeate day-to-day life and have been in the spotlight of mainstream media. If you own a smartphone, for example, you have a thinking camera. We are excited to create a resource that compiles advances of this happening field and help start beginners on their path to understanding.
2. Why was this Specialization created?
Instructors Radhakrishna and Dr. Yuan: Computer vision is a vast interdisciplinary field, with numerous journals and conferences continuously publicizing its advances. Computer vision can be overwhelming for someone who does not know where to begin. This Specialization was created to serve as a primer. Our key focus was on summarizing its evolution and highlighting the perspectives of academia and industry, as both are big players in advancing this research area.
Instructors Radhakrishna and Dr. Yuan: The ideal learner is anyone who wants to learn computer vision fundamentals, such as an undergraduate/graduate student or professional working in a STEM discipline. The first course gives an overview of concepts and applications and the next three cover them in detail. Learners should have some programming experience (in any computer language) and basic knowledge in mathematics, especially linear algebra, calculus and probability theory.
4. What can someone expect to learn?
Instructors Radhakrishna and Dr. Yuan: After successful completion, learners will be well versed in computer vision concepts and key application areas. Through four projects, they will become very familiar with programming in MATLAB®, and acquire confidence in implementing new image processing, computer vision and machine learning projects. They will also be aware of venues/resources to monitor in order to keep pace with this rapidly evolving field.
5. What is MATLAB?
Dr. Armstrong: MATLAB® is the leading mathematical computing software for engineers and scientists developed by Mathworks. People are often surprised to learn that MATLAB is nearly 40 years old. Cleve Moler, professor of math and computer science, created MATLAB using syntax that mirrors common science and engineering notation, so his students could focus on solving math and engineering problems. Today, over 3 million people around the world use MATLAB to solve complex problems in industries such as aerospace, automotive, and energy production.
Dr. Armstrong: Students will learn, apply, and build skills in computer vision using an industry standard tool. They’ll first learn image processing theory and then dive into real world examples. For example, self-driving cars require computers to detect image features like lanes, track the motion of objects such as cars, and recognize items like pedestrians. Learners will use MATLAB to implement these fundamental concepts through projects in each course.
Students will also get to use MATLAB apps. These are built-in graphical tools that enable rapid prototyping and allow for fast experimentation of ideas without writing code, so learners can focus on key concepts.
7. What is particularly exciting about this new Specialization?
Instructors Radhakrishna and Dr. Yuan: Despite recent advancements, there are still many unsolved problems in the computer vision field and there is no fixed way to solve them. The same problem has been approached with different solutions. That is why the fourth course – which introduces deep learning – is very exciting! We demonstrate how several problems discussed in earlier courses are solved effectively by using deep learning. But does that mean deep learning can solve all computer vision issues? We gather various perspectives from industry and academia to answer this question.
Dr. Armstrong: Computer vision is an incredibly important field with applications from autonomous robots to cancer detection. Solving these challenges requires working with large cross-disciplinary teams. MATLAB gives scientists and engineers from many different fields the ability to work together to accomplish tasks – from simulation and prototyping to deploying code on actual devices – more quickly than traditional coding. Students completing this Specialization will be more equipped for a career in computer vision and gain valuable experience with MATLAB, an in-demand software package that’s a required skill for many jobs in this area.
MathWorks is excited to partner with SUNY to give learners access to the same tools used by computer vision professionals. As part of their MOOC Support Program, MathWorks collaborates with Coursera partners to provide free support using their software. They work one-on-one with instructors to create auto-graded assignments, give access to MATLAB, and provide learning resources that teach software functionality. For Coursera partners interested in learning more, email email@example.com.