Over two millennia ago, the ancient mathematician Euclid, widely recognized as the father of geometry, shifted our perspective on shapes. Today, Justin Solomon of MIT uses contemporary geometric methods to tackle complex challenges seemingly unrelated to shapes. Solomon utilizes geometric tools to analyze high-dimensional datasets, providing insights about the potential performance of machine learning models. As an associate professor in the MIT Department of Electrical Engineering and Computer Science and a member of the Computer Science and Artificial Intelligence Laboratory (CSAIL), Solomon argues that using geometry in data analysis is beneficial due to its focus on distances, similarities, curvature, and shape.
Solomon leads the Geometric Data Processing Group, which splits its research focus between processing 2D and 3D data and conducting high-dimensional statistical analysis. They engage in tasks such as aligning 3D organ scans for medical imaging, helping autonomous vehicles recognize pedestrians through spatial data from LiDAR sensors, and developing generative AI models. Solomon asserts the algorithms they’ve created for computer animation also work well for tasks in generative AI and statistics.
Solomon’s journey in this field began with an early interest in computer graphics, leading to internships at a research lab and Pixar Animation Studios. These experiences led him to double major in mathematics and computer science at Stanford University. At Pixar, he worked on physical simulation and rendering techniques for improving the realism of animated films. After deciding to pursue academia, Solomon earned his PhD at Stanford, focusing on optimal transport, a mathematical problem concerning efficiency of distribution.
At MIT, Solomon strives to make geometric research accessible, especially for students who often lack the opportunity to do research. He founded the Summer Geometry Initiative, a program mostly for students from underrepresented backgrounds that offers a paid introduction to geometry research. Since the program’s inception, Solomon has noticed a positive shift in the diversity of incoming PhD students at MIT and other institutions.
Solomon is also eager to use geometric tools to refine unsupervised machine learning models, which learn to recognize patterns without labeled training data. This could involve interpreting complex, unlabeled 3D scenes to improve learning efficiency.
Outside of research, Solomon is passionate about music. An accomplished pianist and cellist, he enjoys performing with local symphonies. He finds the analytical nature of music complementary to his research field, enhancing both his academic and artistic pursuits.