Papers
6
Total Citations
124
H-Index
5
About
Haider Ali is a computer vision and robotics researcher whose work bridges the critical gap between 2D perception and 3D understanding. His primary research areas include 3D pose estimation, articulated robot kinematics, and autonomous navigation. Ali made significant contributions to 3D pose estimation from single 2D images—a fundamental challenge for autonomous driving, robot manipulation, and augmented reality. His most influential work, "3D Pose Regression Using Convolutional Neural Networks" (2017, 67 citations), pioneered a regression-based approach that directly predicts continuous 3D pose parameters, moving beyond the traditional discretized classification methods. He further advanced this paradigm with his mixed classification-regression framework (2018, 24 citations), demonstrating that hybrid approaches can leverage the strengths of both formulations. Beyond pose estimation, Ali has contributed to robotics calibration, developing methods for joint origin identification using multi-camera tracking systems, and to autonomous navigation through his work on road traversability analysis. His research on appearance learning for satellite pose detection at close-range (2017, 10 citations) showcases the breadth of his applications, extending from terrestrial robotics to space operations.
Research Focus
Key Achievements
Top Papers
- 13D Pose Regression Using Convolutional Neural Networks67 citations · 2017
- 2
- 3
- 4Appearance learning for 3D pose detection of a satellite at close-range10 citations · 2017
- 5
- 6Road Traversability analysis using network properties of roadmaps5 citations · 2016