Papers
8
Total Citations
222
H-Index
6
About
Jiayi Ma is a leading researcher in robotics and computer vision, whose work centers on point set registration, loop closure detection (LCD), and visual navigation for autonomous systems. His major contributions lie in developing robust, learning-based methods for nonrigid point set registration, where his 2018 paper on robust transformation learning under manifold regularization (148 citations) has become a foundational reference in the field. Ma has also made significant advances in simultaneous localization and mapping (SLAM), particularly through appearance-based LCD techniques that leverage locality-driven motion field learning and bidirectional manifold representation consensus to improve robot localization accuracy. His work on topological navigation using convolutional neural network features and sharpness measures has further advanced scalable map representations for large-scale environments. With multiple papers on LCD and visual place recognition, Ma’s research consistently addresses the challenge of drift in robot pose estimation, achieving high impact through innovative geometric matching and consensus algorithms. His publications, including those in top venues, demonstrate a sustained focus on making autonomous navigation more reliable and efficient, earning him recognition as a key contributor to modern robotic perception systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8Visual Place Recognition via Local Affine Preserving Matching2 citations · 2021