Deep Measurement Updates for Bayes Filters

TL;DR

This paper introduces Deep Measurement Update (DMU), a neural network-based method for measurement updates in Bayes filters that effectively processes high-dimensional sensor data like images, trained on synthetic data and validated on real-world scenarios.

Contribution

The paper presents DMU, a novel neural network approach for measurement updates in Bayes filters, capable of handling high-dimensional data without hand-crafted heuristics, trained efficiently with primed data scheme.

Findings

DMU performs well on real-world data despite training only on synthetic data.

The method effectively estimates pose and internal states in complex articulated systems.

Benchmark comparison shows DMU's competitive performance against A-SDF on RBO dataset.

Abstract

Measurement update rules for Bayes filters often contain hand-crafted heuristics to compute observation probabilities for high-dimensional sensor data, like images. In this work, we propose the novel approach Deep Measurement Update (DMU) as a general update rule for a wide range of systems. DMU has a conditional encoder-decoder neural network structure to process depth images as raw inputs. Even though the network is trained only on synthetic data, the model shows good performance at evaluation time on real-world data. With our proposed training scheme primed data training , we demonstrate how the DMU models can be trained efficiently to be sensitive to condition variables without having to rely on a stochastic information bottleneck. We validate the proposed methods in multiple scenarios of increasing complexity, beginning with the pose estimation of a single object to the joint estimation of the pose and the internal state of an articulated system. Moreover, we provide a benchmark against Articulated Signed Distance Functions(A-SDF) on the RBO dataset as a baseline comparison for articulation state estimation.