Mohammad Khodadad

Sharif University of Technology, McMaster University

Papers

Total Citations

H-Index

About

Mohammad Khodadad is a rising researcher in 3D shape analysis and geometric deep learning, whose work bridges theoretical foundations with practical efficiency. His key contributions center on developing novel mathematical representations and ultra-efficient neural architectures for processing 3D point clouds and surface meshes. In his highly cited 2022 work, "Medial Spectral Coordinates for 3D Shape Analysis," Khodadad introduced a powerful new coordinate system that leverages medial axis information to enable robust shape matching and analysis, addressing long-standing challenges in handling noisy or incomplete 3D data. Building on this, his 2024 paper "MLGCN: an ultra efficient graph convolutional neural model for 3D point cloud analysis" presents a groundbreaking lightweight architecture that dramatically reduces computational cost while maintaining high accuracy on classification and segmentation tasks—a critical advance for real-time applications on resource-constrained devices. With his work already accumulating citations from leading computer vision and graphics venues, Khodadad is establishing himself as a key innovator in making 3D analysis both theoretically elegant and practically deployable, promising to accelerate the adoption of 3D sensing technologies across robotics, autonomous systems, and augmented reality.