StaticNeRF: Neural Implicit Static Mapping and Localization in Dynamic Environments

Lee Juhui, Seungjun Ma, Geonmo Yang, Younggun Cho

Keywords: NeRF, Localization, Robotics

TL;DR: Neural Implicit Static Mapping and Localization in Dynamic Environments

Abstract: Recently, neural implicit representations have been widely introduced for robot mapping to achieve highresolution maps. Previous approaches perform well in stable, static environments but encounter difficulties when faced with the challenges posed by moving objects. In this paper, we propose a novel neural implicit mapping and robust filter-based localization in dynamic environments. In this paper, we introduce a novel neural implicit mapping and robust filter-based localization in dynamic environments. By incorporating the transient field alongside the static radiance field, the proposed method can construct a static implicit field without any semantic information about dynamic objects. Also, we design the static particle-aware localization pipeline for robust localization in dynamic environments. Our approach is validated against standard and custom datasets, demonstrating that our implicit neural map has better performance than the other neural rendering methods and that it is effective in dynamic object removal and accuracy in localization, marking a step forward for efficient navigation systems.

Submission Number: 28