Backprop KF: Learning Discriminative Deterministic State Estimators

Abstract

Generative state estimators based on probabilistic filters and smoothers are one of the most popular classes of state estimators for robots and autonomous vehicles. However, generative models have limited capacity to handle rich sensory observations, such as camera images, since they must model the entire distribution over sensor readings. Discriminative models do not suffer from this limitation, but are typically more complex to train as latent variable models for state estimation. We present an alternative approach where the parameters of the latent state distribution are directly optimized as a deterministic computation graph, resulting in a simple and effective gradient descent algorithm for training discriminative state estimators. We show that this procedure can be used to train state estimators that use complex input, such as raw camera images, which must be processed using expressive nonlinear function approximators such as convolutional neural networks. Our model can be viewed as a type of recurrent neural network, and the connection to probabilistic filtering allows us to design a network architecture that is particularly well suited for state estimation. We evaluate our approach on synthetic tracking task with raw image inputs and on the visual odometry task in the KITTI dataset. The results show significant improvement over both standard generative approaches and regular recurrent neural networks.

NIPS 2016

References in corpus (1)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

Cited by in corpus (17)

• KalmanNet: Neural Network Aided Kalman Filtering for Partially Known Dynamics
• A Disentangled Recognition and Nonlinear Dynamics Model for Unsupervised Learning
• Robust Data-Driven Zero-Velocity Detection for Foot-Mounted Inertial Navigation
• DPC-Net: Deep Pose Correction for Visual Localization
• Adaptive Kalman-Informed Transformer
• PVEs: Position-Velocity Encoders for Unsupervised Learning of Structured State Representations
• QMDP-Net: Deep Learning for Planning under Partial Observability
• How to Train a CAT: Learning Canonical Appearance Transformations for Direct Visual Localization Under Illumination Change
• Self-Supervised Deep Pose Corrections for Robust Visual Odometry
• Factor Graph-Based Smoothing Without Matrix Inversion for Highly Precise Localization
• Recurrent Predictive State Policy Networks
• Interpretable Deep Feature Propagation for Early Action Recognition
• Deep execution monitor for robot assistive tasks
• Particle Filter Recurrent Neural Networks
• Integrating Algorithmic Planning and Deep Learning for Partially Observable Navigation
• Estimating Nonlinear Dynamics with the ConvNet Smoother
• Online Visual Robot Tracking and Identification using Deep LSTM Networks