Manohar Kaul

Indian Institute of Technology Hyderabad

Papers

Total Citations

H-Index

About

Manohar Kaul is a prominent researcher specializing in 3D point cloud learning, self-supervised learning, and domain adaptation, with a particular focus on advancing machine learning techniques for geometric and spatial data. His work sits at the intersection of deep learning and 3D computer vision, addressing challenges critical to real-world applications in robotics, autonomous driving, and virtual reality. Kaul's most notable contribution, "Self-Supervised Few-Shot Learning on Point Clouds" (2020), tackles the costly bottleneck of labeled data in 3D deep learning by leveraging self-supervised frameworks to enable effective learning from minimal annotations — a breakthrough that has garnered 47 citations and influenced subsequent research in data-efficient 3D learning. His more recent work on synergizing contrastive learning with optimal transport for point cloud domain adaptation (2024) demonstrates his continued push toward robust, label-efficient 3D representation learning, addressing the persistent challenge of domain shift in real-world deployments. Collectively, Kaul's research reflects a coherent and forward-thinking agenda: making deep learning on 3D data more practical, generalizable, and accessible. His contributions are increasingly relevant as industries scale deployment of LiDAR-based systems and 3D scene understanding technologies.