Using the principles of geometry, Justin Solomon, an associate professor in MIT’s Department of Electrical Engineering and Computer Science, is tackling complex problems in data science and computer graphics. Building on Euclid’s ancient foundations of geometry, Solomon is leveraging geometric techniques to solve problems that are seemingly unrelated to shapes. He asserts that the language we use to discuss data often involves distances, similarities, and shapes, aspects that have been integral to geometry.
Among his projects, he is using geometric techniques for comparing datasets for machine learning models. Furthermore, his team grapples with tasks that involve the processing of two- and three-dimensional geometric data. This includes aligning 3D organ scans in medical imaging and facilitating autonomous vehicles in identifying pedestrians using spatial data collected by LiDAR sensors.
The remaining team members conduct high-dimensional statistical research using geometric techniques. Such research is employed to develop better generative AI models that construct new images from specific parts of an example image-filled dataset. According to Solomon, mapping the space of images is fundamentally a geometric problem.
Solomon’s curiosity in computer graphics began early on. This interest steered him to major in mathematics and computer science at Stanford University. He would later work at Pixar Animation Studios as a summer intern and eventually graduate school student, focusing on the physical simulation of cloth and fluids, which enhanced the realism of animated movies.
He remained at Stanford to pursue his Ph.D. in computer science, studying the problem of optimal transport. He researched how to transfer a distribution of goods to another distribution as efficiently as possible. Though his initial focus was dedicated to computer graphic applications, his research extended to different directions and practices which shaped the structure of his research group at MIT.
To make geometric research accessible, Solomon initiated the Summer Geometry Initiative. It’s a six-week research program for undergraduates, primarily from underrepresented backgrounds. He aspires to change the status quo of the typical candidate coming from a narrow selection of schools and extends opportunities to those who may not have been privileged earlier. Solomon hopes to apply geometric tools to refine unsupervised machine learning models in the future.
Outside his research responsibilities, Solomon maintains a keen interest in classical music. He greatly appreciates composer Dmitri Shostakovich and participates in a symphony in every city he moves to, presently playing the cello with the New Philharmonia Orchestra in Newton, Massachusetts.
For Solomon, music is analytical in nature and is complemented by his research field, computer graphics, which is closely related to artistic practice. Solomon believes that music and graphics are mutually beneficial, reflecting his seamless integration of varied interests in his work and life.
