Simultaneous Localization and Mapping (SLAM) technology has remarkably progressed over the last two decades, and it is critical for the autonomous exploration of unknown environments. This evolution has spanned from manually-crafted methods through the deep learning era to the recent implementation of Neural Radiance Fields (NeRFs) and 3D Gaussian Splatting (3DGS).

• Overview: A comprehensive survey on SLAM’s progress via NeRFs and 3DGS.
• Evolution: Discusses SLAM’s transition from hand-crafted methods to deep learning, before the arrival of NeRFs and 3DGS.
• Significance: Highlights strengths, limitations, and serves as a reference for ongoing dynamic progress and challenges in the field.
• Background: Explains the technological background and its importance to autonomous systems.

This paper’s exploration of SLAM’s advancements through NeRFs and 3DGS provides an insightful look into the trajectory of autonomous navigation technologies. The focus on emerging techniques underscores the necessity for constant innovation in the AI field and the potential these methods have in further refining and enhancing systems that can understand and interact with their environments.