Advancements in Embodied Navigation and SLAM Systems

The field of embodied navigation and SLAM systems is moving towards more robust and efficient methods for navigating complex environments. Researchers are exploring new approaches to improve the accuracy and reliability of these systems, including the use of 3D Gaussian Splatting, dynamic-aware LIO frameworks, and multi-agent cooperative SLAM. Notable papers in this area include Towards Physically Executable 3D Gaussian for Embodied Navigation, which proposes a new paradigm for upgrading 3DGS into an executable, semantically and physically aligned environment. Another noteworthy paper is LT-Exosense, which introduces a vision-centric, multi-session mapping system for long-term autonomous navigation of exoskeletons. Additionally, Breaking the Static Assumption presents a dynamic-aware LIO framework that integrates dynamic awareness directly into the point cloud registration process, and RoGER-SLAM proposes a robust 3DGS SLAM system tailored for noise and low-light resilience. These advancements have the potential to significantly improve the performance of embodied navigation and SLAM systems in various applications.

Sources

Towards Physically Executable 3D Gaussian for Embodied Navigation

LT-Exosense: A Vision-centric Multi-session Mapping System for Lifelong Safe Navigation of Exoskeletons

Breaking the Static Assumption: A Dynamic-Aware LIO Framework Via Spatio-Temporal Normal Analysis

Bag-of-Word-Groups (BoWG): A Robust and Efficient Loop Closure Detection Method Under Perceptual Aliasing

RoGER-SLAM: A Robust Gaussian Splatting SLAM System for Noisy and Low-light Environment Resilience

LVD-GS: Gaussian Splatting SLAM for Dynamic Scenes via Hierarchical Explicit-Implicit Representation Collaboration Rendering

TWC-SLAM: Multi-Agent Cooperative SLAM with Text Semantics and WiFi Features Integration for Similar Indoor Environments

An approach for combining transparency and motion assistance of a lower body exoskeleton

AgriGS-SLAM: Orchard Mapping Across Seasons via Multi-View Gaussian Splatting SLAM

Built with on top of