Robust Gaussian filtering using a pseudo measurement

Wüthrich, Manuel; Cifuentes, Cristina Garcia; Trimpe, Sebastian; Meier, Franziska; Bohg, Jeannette; Issac, Jan; Schaal, Stefan

doi:10.1109/acc.2016.7525473

Cited by 11 publications

(7 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ideas in this paper are not intended to solve one concrete filtering problem, but rather to provide a theoretical basis for future research. In fact, these insights have already given rise to practical filtering algorithms in Wüthrich et al (2016) and Issac et al (2016). The first reference proposes using a measurement feature for robustifying GFs against outliers.…”

Section: Discussionmentioning

confidence: 99%

“…The last two examples illustrate how features can be designed for specific systems. Another example of a designed feature is given in Wüthrich et al (2016), where the authors derive a feature in order to handle fat-tailed measurement models.…”

Section: Simulation Examples Illustrating the Benefit Of Measuremementioning

confidence: 99%

“…While no full-scale robotic filtering example is presented herein, the results of this paper have already given rise to practical applications. In Wüthrich et al (2016), a method for robustifying GFs against outliers using a pseudo measurement is presented, which is applied to three-dimensional (3D) object tracking using a depth sensor in Issac et al (2016).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A new perspective and extension of the Gaussian Filter

Wüthrich

Trimpe

Cifuentes

et al. 2016

The International Journal of Robotics Research

Self Cite

View full text Add to dashboard Cite

The Gaussian Filter (GF) is one of the most widely used filtering algorithms; instances are the Extended Kalman Filter, the Unscented Kalman Filter and the Divided Difference Filter. The GF represents the belief of the current state by a Gaussian distribution, whose mean is an affine function of the measurement. We show that this representation can be too restrictive to accurately capture the dependences in systems with nonlinear observation models, and we investigate how the GF can be generalized to alleviate this problem. To this end, we view the GF as the solution to a constrained optimization problem. From this new perspective, the GF is seen as a special case of a much broader class of filters, obtained by relaxing the constraint on the form of the approximate posterior. On this basis, we outline some conditions which potential generalizations have to satisfy in order to maintain the computational efficiency of the GF. We propose one concrete generalization which corresponds to the standard GF using a pseudo measurement instead of the actual measurement. Extending an existing GF implementation in this manner is trivial. Nevertheless, we show that this small change can have a major impact on the estimation accuracy.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Simulation Examples Illustrating the Benefit Of Measuremementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A new perspective and extension of the Gaussian Filter

Wüthrich

Trimpe

Cifuentes

et al. 2016

The International Journal of Robotics Research

Self Cite

View full text Add to dashboard Cite

show abstract

“…At present, the modeling method of HTMN is mainly realized through the combination of Gaussian noise and other noise distributions. The HTMN was described as a weighted sum of two Gaussian distributions in [22]- [26]. One Laplace distribution and one small variance Gaussian distribution were employed to characterize the HTMN together in [21], [27] and [28].…”

Section: Introductionmentioning

confidence: 99%

Robust Generalized Labeled Multi-Bernoulli Filter for Multitarget Tracking With Unknown Non- Stationary Heavy-Tailed Measurement Noise

et al. 2021

View full text Add to dashboard Cite

A robust generalized labeled multi-Bernoulli (GLMB) filter is presented to perform multitarget tracking (MTT) with unknown non-stationary heavy-tailed measurement noise (HTMN). The HTMN is modeled as a multivariate Student's t-distribution with unknown and time-varying mean. The proposed filter relaxes the restrictive assumption that the mean of HTMN is zero, and can effectively deal with MTT under the condition that the mean of HTMN is unknown and time-varying. The variational Bayesian (VB) approximation is applied in the GLMB filtering framework with the augmented state. The marginal likelihood function is obtained via minimizing the Kullback-Leibler divergence by the variational lower bound. The simulation results demonstrate that the proposed filter can effectively track multiple targets in both linear and nonlinear scenarios when the mean of HTMN is unknown and time-varying.INDEX TERMS Generalized labeled multi-Bernoulli filter, multitarget tracking (MTT), variational Bayesian (VB), non-stationary, heavy-tailed measurement noise (HTMN), unknown and time-varying mean.

show abstract

“…This is indeed the case in deep learning as well as in many Bayesian inference techniques. While there are many methods to adapt deep neural networks to varying domains [1], [2], [3], [4], [5], such adaptation techniques are under-explored for Bayesian models [6] despite their extensive applications in robotics [7], [8], [9], [10], [11]. As uncertainty is represented as probability distributions in Bayesian models, entire distributions need to be adapted when changing to a new domain.…”

Section: Introductionmentioning

confidence: 99%

Online Domain Adaptation for Occupancy Mapping

Tompkins¹,

Senanayake²,

Ramos³

2020

Robotics: Science and Systems XVI

View full text Add to dashboard Cite

Creating accurate spatial representations that take into account uncertainty is critical for autonomous robots to safely navigate in unstructured environments. Although recent LIDAR based mapping techniques can produce robust occupancy maps, learning the parameters of such models demand considerable computational time, discouraging them from being used in real-time and large-scale applications such as autonomous driving. Recognizing the fact that realworld structures exhibit similar geometric features across a variety of urban environments, in this paper, we argue that it is redundant to learn all geometry dependent parameters from scratch. Instead, we propose a theoretical framework building upon the theory of optimal transport to adapt model parameters to account for changes in the environment, significantly amortizing the training cost. Further, with the use of high-fidelity driving simulators and real-world datasets, we demonstrate how parameters of 2D and 3D occupancy maps can be automatically adapted to accord with local spatial changes. We validate various domain adaptation paradigms through a series of experiments, ranging from inter-domain feature transfer to simulation-to-real-world feature transfer. Experiments verified the possibility of estimating parameters with a negligible computational and memory cost, enabling large-scale probabilistic mapping in urban environments.

show abstract

Robust Gaussian filtering using a pseudo measurement

Cited by 11 publications

References 32 publications

A new perspective and extension of the Gaussian Filter

A new perspective and extension of the Gaussian Filter

Robust Generalized Labeled Multi-Bernoulli Filter for Multitarget Tracking With Unknown Non- Stationary Heavy-Tailed Measurement Noise

Online Domain Adaptation for Occupancy Mapping

Contact Info

Product

Resources

About