A Kalman filter for robust outlier detection

— In this paper, we introduce a modified Kalman filter that can perform robust, real-time outlier detection in the observations, without the need for parameter tuning. Robotic systems that rely on high quality sensory data can be sensitive to data containing outliers. Since the standard Kalman filter is not robust to outliers, other variations of the Kalman filter have been proposed to overcome this issue, but these methods may require parameter tuning, use of heuristics or complicated parameter estimation. Our Kalman filter uses a weighted least squares-like approach by introducing weights for each data sample. A data sample with a smaller weight has a weaker contribution when estimating the current time step’s state. We learn the weights and system dynamics using a variational Expectation-Maximization framework. We evaluate our Kalman filter algorithm on synthetic data and data from a robotic dog.

• Type: Conference
• Year: 2007
• Where: IROS
• Authors: Jo-Anne Ting, Evangelos Theodorou, Stefan Schaal
• Keywords: IROS 2007, Kalman filter, Modified Kalman filter, Parameter Tuning, Robotics