Max Robotics — Global Service Robot Directory + US Certification

About

Ameesh Makadia is a leading researcher in computer vision and robotics, whose work fundamentally advances how machines perceive and navigate 3D space. His primary research areas include rotation estimation, ego-motion analysis, and 3D geometry, with a particular focus on developing robust, correspondenceless methods for spatial reasoning. Makadia’s seminal contributions began with his pioneering work on direct 3D-rotation estimation from spherical images, where he introduced a generalized shift theorem that enables appearance-based robot localization without explicit feature matching—a paper that has garnered 62 citations and remains foundational in omnidirectional vision. He further extended these ideas to planar and IMU-assisted ego-motion estimation, demonstrating how constrained camera motions and sensor fusion can simplify localization tasks. More recently, Makadia has made significant strides in deep learning for rotation estimation, notably with his analysis of SVD for symmetric orthogonalization (32 citations) and the introduction of Implicit-PDF, a non-parametric representation for probability distributions on the rotation manifold that addresses uncertainty and symmetry in single-image pose estimation. His work on sum-of-squares polynomials for 3D environment geometry also showcases his versatility in applying algebraic methods to spatial reasoning. With over 160 combined citations across his most influential papers, Makadia’s research continues to shape how robots and autonomous systems understand and interact with their surroundings.