About Me
My CV is an "official" description of things I have done as a student that are relevant to my current studies. There are many constraints on such a document, so this is meant to be a more detailed and freeform description of the things I do in life. This is completely about me and what I think about myself. It is probably very selective and biased. If you care enough to read it, then take it with a grain of salt and a healthy skepticism for the things that one writes about oneself.
What I study
I am a student of figuring out how cool things work. Initially, that was anything mechanical (I used to take everything apart was I was really litte) but now it is more complex stuff, like thought and vision. I am a Computer Science concentrator at Harvard College (which is the undergraduate school at Harvard University, for those who are confused by the distinction.) I am on the Mind Brain and Behavior track within that concentration, which means I am trying to learn things that tie computer science into the larger study of how humanity thinks.
Ever since I took a class on it in the fall, I've become particularly interested in Computer Vision. Computer Vision is best described as an ill-posed inverse problem, as Prof. Todd Zickler described it to me in his class on the subject. The world is full of objects that reflect light. Those objects are illuminated by light sources that provide different kinds of light. By combining the direction to and distance to and object, the relative position of the light source, the reflective qualities of the object, and the quality of the light source, we can compute an image that is a function of all of these qualities. Vision asks us to invert this function and determine from the image what features went into creating it.
The reason that this is ill-posed is that there are an infinite number of possible configurations of the world that would yield the same image, so there is no clear inverse. Despite this, our brains are perfectly capable of decerning accurate information from images in almost all cases, so somehow we are able to narrow down the problem to reasonable candidates. This is the challenge of computer vision. We must take a problem that evolution has already solved and determine how we can find our own solution using computers and algorithms. We must narrow down the realm of possibilities so that we can solve with some degree of accuracy this difficult inverse problem.
Solving this problem will lead to many new possibilities. Most obviously, we can use computer vision to perform visual tasks automatically, like recognizing shapes, objects, or even people. We can put vision advances into robotic systems so that they are better able to interact with the world. Less obviously, I believe that each time we solve a problem in computer vision, we learn something about ourselves and our own visual system. Algorithms can help us understand how our own brain solves the say problems, and perhaps they can lead back into neuroscience and help unlock new answers to the brain's mysteries.
Model UN
To be written.
Theater
To be written.
Politics
To be written.
Sailing
To be written.