Incremental Abstraction in Distributed Probabilistic SLAM Graphs

TL;DR

This paper introduces a novel distributed SLAM framework that incrementally builds semantic scene graphs using neural network proposals and Gaussian Belief Propagation for efficient, dense, and compact scene representation in real indoor environments.

Contribution

It presents a new incremental abstraction method integrating neural network proposals into SLAM graphs and a distributed inference technique using Gaussian Belief Propagation for improved efficiency.

Findings

Successfully recovers major planes in indoor scenes

Achieves significant compression of scene representations

Demonstrates speed advantages over direct inference methods

Abstract

Scene graphs represent the key components of a scene in a compact and semantically rich way, but are difficult to build during incremental SLAM operation because of the challenges of robustly identifying abstract scene elements and optimising continually changing, complex graphs. We present a distributed, graph-based SLAM framework for incrementally building scene graphs based on two novel components. First, we propose an incremental abstraction framework in which a neural network proposes abstract scene elements that are incorporated into the factor graph of a feature-based monocular SLAM system. Scene elements are confirmed or rejected through optimisation and incrementally replace the points yielding a more dense, semantic and compact representation. Second, enabled by our novel routing procedure, we use Gaussian Belief Propagation (GBP) for distributed inference on a graph processor. The time per iteration of GBP is structure-agnostic and we demonstrate the speed advantages over direct methods for inference of heterogeneous factor graphs. We run our system on real indoor datasets using planar abstractions and recover the major planes with significant compression.