Simultaneous Latent State Estimation and Latent Linear Dynamics Discovery from Image Observations

TL;DR

This paper reviews existing methods for latent state estimation from image observations and proposes a new approach that combines latent state estimation with linear dynamics discovery.

Contribution

It introduces a novel method that simultaneously estimates latent states and discovers linear dynamics from image data, advancing prior separate or sequential approaches.

Findings

Demonstrates improved accuracy in latent state estimation from images.

Shows effective discovery of underlying linear dynamics.

Provides a unified framework for state estimation and dynamics learning.

Abstract

The problem of state estimation has a long history with many successful algorithms that allow analytical derivation or approximation of posterior filtering distribution given the noisy observations. This report tries to conclude previous works to resolve the problem of latent state estimation given image-based observations and also suggests a new solution to this problem.