Incremental Non-Gaussian Inference for SLAM Using Normalizing Flows
This addresses the challenge of non-Gaussian inference in SLAM for robotics, offering incremental updates similar to iSAM2 but with improved accuracy, though it is incremental/hybrid in method.
The paper tackles the problem of inferring full posterior distributions in SLAM with non-Gaussian factors by introducing NF-iSAM, which uses normalizing flows and neural networks to model and sample the posterior, showing superior accuracy in describing beliefs for continuous and discrete variables in range-only SLAM.
This paper presents normalizing flows for incremental smoothing and mapping (NF-iSAM), a novel algorithm for inferring the full posterior distribution in SLAM problems with nonlinear measurement models and non-Gaussian factors. NF-iSAM exploits the expressive power of neural networks, and trains normalizing flows to model and sample the full posterior. By leveraging the Bayes tree, NF-iSAM enables efficient incremental updates similar to iSAM2, albeit in the more challenging non-Gaussian setting. We demonstrate the advantages of NF-iSAM over state-of-the-art point and distribution estimation algorithms using range-only SLAM problems with data association ambiguity. NF-iSAM presents superior accuracy in describing the posterior beliefs of continuous variables (e.g., position) and discrete variables (e.g., data association).