Abstract:Meticulous 3D environment representations have been a longstanding goal in computer vision and robotics fields. The recent emergence of neural implicit representations has introduced radical innovation to this field as implicit representations enable numerous capabilities. Among these, the Neural Radiance Field (NeRF) has sparked a trend because of the huge representational advantages, such as simplified mathematical models, compact environment storage, and continuous scene representations. Apart from computer vision, NeRF has also shown tremendous potential in the field of robotics. Thus, we create this survey to provide a comprehensive understanding of NeRF in the field of robotics. By exploring the advantages and limitations of NeRF, as well as its current applications and future potential, we hope to shed light on this promising area of research. Our survey is divided into two main sections: \textit{The Application of NeRF in Robotics} and \textit{The Advance of NeRF in Robotics}, from the perspective of how NeRF enters the field of robotics. In the first section, we introduce and analyze some works that have been or could be used in the field of robotics from the perception and interaction perspectives. In the second section, we show some works related to improving NeRF's own properties, which are essential for deploying NeRF in the field of robotics. In the discussion section of the review, we summarize the existing challenges and provide some valuable future research directions for reference.

Autonerf: Training implicit scene representations with autonomous agents

AutoNeRF: Training Implicit Scene Representations with Autonomous Agents

Towards Embodied Neural Radiance Fields

Benchmarking Neural Radiance Fields for Autonomous Robots: An Overview

Active Perception using Neural Radiance Fields

Finding Waldo: Towards Efficient Exploration of NeRF Scene Spaces

Reconstructive Latent-Space Neural Radiance Fields for Efficient 3D Scene Representations

ActiveNeRF: Learning Where to See with Uncertainty Estimation

Multi-Object Navigation with dynamically learned neural implicit representations

NeurAR: Neural Uncertainty for Autonomous 3D Reconstruction With Implicit Neural Representations

Autonomous Implicit Indoor Scene Reconstruction with Frontier Exploration

NeRF: Neural Radiance Field in 3D Vision, A Comprehensive Review

Learning autonomous driving from aerial imagery

Integrating Neural Radiance Fields with Deep Reinforcement Learning for Autonomous Mapping of Small Bodies

NeuRAD: Neural Rendering for Autonomous Driving

NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections

Integrating Neural Radiance Fields End-to-End for Cognitive Visuomotor Navigation

One-Shot Neural Fields for 3D Object Understanding

NeRF in Robotics: A Survey

Neural Radiance Field in Autonomous Driving: A Survey