Jo-Anne Ting

About

Jo-Anne Ting is a robotics and machine learning researcher whose work sits at the intersection of probabilistic methods, robust estimation, and autonomous systems. Her most significant contributions center on developing Bayesian and statistical techniques to address real-world challenges in robotic sensing and control. Most notably, her 2007 work introducing a robust Kalman filter for outlier detection — cited 143 times — provided a practical, parameter-free solution to a persistent problem in robotic sensory systems, where unreliable data can critically undermine autonomous operation. Complementing this, her Bayesian approach to nonlinear parameter identification for rigid body dynamics (61 citations) offered a principled framework for modeling complex humanoid robots where traditional CAD-based methods fall short. Ting's research extends into tactile sensing, where she explored active estimation of object dynamics and sequential learning with tactile feedback, advancing dexterous robotic manipulation. Her work on high-dimensional regression and kernel shaping further demonstrates a breadth spanning both theoretical machine learning and applied robotics. Across her career, Ting has consistently championed automation of model tuning and uncertainty quantification, making sophisticated robotic systems more reliable and accessible without demanding extensive manual calibration from practitioners.

Research Focus

Key Achievements

Top Papers

1. 1
   A Kalman filter for robust outlier detection

   143 citations · 2007
2. 2
   A Bayesian Approach to Nonlinear Parameter Identification for Rigid Body Dynamics

   61 citations · 2006
3. 3
   Learning an Outlier-Robust Kalman Filter

   56 citations · 2007
4. 4
   Automatic Outlier Detection: A Bayesian Approach

   44 citations · 2007
5. 5
   Active estimation of object dynamics parameters with tactile sensors

   42 citations · 2010
6. 6
   Bayesian robot system identification with input and output noise

   29 citations · 2010
7. 7
   Efficient Learning and Feature Selection in High-Dimensional Regression

   26 citations · 2009
8. 8
   Active Sequential Learning with Tactile Feedback

   17 citations · 2010
9. 9
   A Bayesian approach to empirical local linearization for robotics

   16 citations · 2008
10. 10
    Bayesian Kernel Shaping for Learning Control

    14 citations · 2008

Key Collaborators

Contact & Links