CELLO: A fast algorithm for Covariance Estimation

We present CELLO (Covariance Estimation and Learning through Likelihood Optimization), an algorithm for predicting the covariances of measurements based on any available informative features. This algorithm is intended to improve the accuracy and reliability of on-line state estimation by providing a principled way to extend the conventional fixed-covariance Gaussian measurement model. We show that in experiments, CELLO learns to predict measurement covariances that agree with empirical covariances obtained by manually annotating sensor regimes. We also show that using the learned covariances during filtering provides substantial quantitative improvement to the overall state estimate.

Authors

• William Vega-Brown
• Abraham Bachrach
• Adam Bry
• Jonathan Kelly
• Nicholas Roy

Keywords

• Vectors
• Prediction algorithms
• Kalman filters
• Measurement
• Manganese
• Estimation
• Robot sensing systems
• Covariance Estimation
• Measurement Covariance
• Principled Way
• Empirical Covariance
• Likelihood Optimization
• Matrix Elements
• Kalman Filter
• Weight Function
• Gaussian Process
• Positive Function
• Positive Definite Matrix
• Element Of Vector
• Optical Flow
• Body Of Data
• Output Of Algorithm
• Error Vector
• Intuitive Way
• Nearest Neighbor Search
• Motion Estimation
• Scale Matrix
• Unscented Kalman Filter
• Optimal Mapping
• Prediction Vector
• Predictor Space
• Outer Product
• Learning Process
• Measurement Noise Covariance
• Latent State
• Symmetric Positive Definite Matrix
• Motion Capture