Greek mathematician Euclid revolutionized the concept of shapes over two millennia ago, laying a strong foundation for geometry. Justin Solomon, leveraging his ancient principles with modern geometric techniques, is solving complex issues unrelated to shapes.
Solomon, an associate professor at MIT Department of Electrical Engineering and Computer Science (EECS) and a member of the Computer Science and Artificial Intelligence Laboratory (CSAIL), explains that datasets may share some geometric structure based on their arrangement within high-dimensional space. This perspective helps to understand whether a machine-learning model will work on different datasets. Consequently, geometric tools provide a unique perspective, contributing to the solution of abstract data science problems.
About half of Solomon’s Geometric Data Processing Group focuses on problems involving 2D and 3D geometric data, such as aligning 3D scans in medical imaging or permitting self-driving vehicles to identify pedestrians from the spatial data collected by LiDAR sensors. The rest of the team uses geometric techniques for high-dimensional statistical research, like developing advanced generative AI models.
Solomon’s fascination for computer graphics led to a research-focused academic career, studying physical simulation of cloth, fluids, and rendering techniques at Pixar Animation Studios. His research also concentrated on optimal transport, seeking the most efficient way to move specific items.
Solomon appreciated MIT’s environment, allowing him to work with brilliant students, postdocs, and colleagues on multifaceted but practical issues impacting various fields. He seeks to make the field of geometric research accessible to underserved students, launching a paid summer research program called Summer Geometry Initiative.
Moving forward, Solomon aims to apply geometric tools to enhance unsupervised machine learning models, improving pattern recognition even without labeled training data. With the help of advanced models that incorporate geometric understanding, computers can map complex, unlabeled 3D scenes effectively, facilitating more efficient learning.
Apart from his academic pursuits, Solomon is an avid musician, enjoying classical music, and routinely joining a local symphony in his city of residence. He draws parallels between music and his research, mentioning that both are analytical and contribute to each other, thereby creating a harmonic combination of his interests.
