LATITUDE: Robotic Global Localization with Truncated Dynamic Low-pass Filter in City-scale NeRF

Neural Radiance Fields (NeRFs) have made great success in representing complex 3D scenes with high-resolution details and efficient memory. Nevertheless, current NeRF-based pose estimators have no initial pose prediction and are prone to local optima during optimization. In this paper, we present LATITUDE: Global Localization with Truncated Dynamic Low-pass Filter, which introduces a two-stage localization mechanism in city-scale NeRF. In place recognition stage, we train a regressor through images generated from trained NeRFs, which provides an initial value for global localization. In pose optimization stage, we minimize the residual between the observed image and rendered image by directly optimizing the pose on tangent plane. To avoid convergence to local optimum, we introduce a Truncated Dynamic Low-pass Filter (TDLF) for coarse-to-fine pose registration. We evaluate our method on both synthetic and real-world data and show its potential applications for high-precision navigation in large-scale city scenes. Codes and data will be publicly available at https://github.com/jike5/LATITUDE.

[1]  I. Pratikakis,et al.  A Survey on Map-Based Localization Techniques for Autonomous Vehicles , 2023, IEEE Transactions on Intelligent Vehicles.

[2]  C. Theobalt,et al.  NeuRIS: Neural Reconstruction of Indoor Scenes Using Normal Priors , 2022, ECCV.

[3]  V. Prisacariu,et al.  DFNet: Enhance Absolute Pose Regression with Direct Feature Matching , 2022, ECCV.

[4]  Pratul P. Srinivasan,et al.  Block-NeRF: Scalable Large Scene Neural View Synthesis , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[5]  Martin R. Oswald,et al.  NICE-SLAM: Neural Implicit Scalable Encoding for SLAM , 2021, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[6]  D. Ramanan,et al.  Mega-NeRF: Scalable Construction of Large-Scale NeRFs for Virtual Fly- Throughs , 2021, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[7]  C. Theobalt,et al.  BungeeNeRF: Progressive Neural Radiance Field for Extreme Multi-scale Scene Rendering , 2021, ECCV.

[8]  Pratul P. Srinivasan,et al.  Dense Depth Priors for Neural Radiance Fields from Sparse Input Views , 2021, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[9]  Jeannette Bohg,et al.  Vision-Only Robot Navigation in a Neural Radiance World , 2021, IEEE Robotics and Automation Letters.

[10]  Hui Huang,et al.  Capturing, Reconstructing, and Simulating: The UrbanScene3D Dataset , 2021, ECCV.

[11]  D. Ramanan,et al.  Depth-supervised NeRF: Fewer Views and Faster Training for Free , 2021, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[12]  Cédric Demonceaux,et al.  FLYBO: A Unified Benchmark Environment for Autonomous Flying Robots , 2021, 2021 International Conference on 3D Vision (3DV).

[13]  Arnaud de La Fortelle,et al.  LENS: Localization enhanced by NeRF synthesis , 2021, CoRL.

[14]  Antonio Torralba,et al.  BARF: Bundle-Adjusting Neural Radiance Fields , 2021, 2021 IEEE/CVF International Conference on Computer Vision (ICCV).

[15]  Zirui Wang,et al.  Direct-PoseNet: Absolute Pose Regression with Photometric Consistency , 2021, 2021 International Conference on 3D Vision (3DV).

[16]  Edgar Sucar,et al.  iMAP: Implicit Mapping and Positioning in Real-Time , 2021, 2021 IEEE/CVF International Conference on Computer Vision (ICCV).

[17]  M. Pollefeys,et al.  Back to the Feature: Learning Robust Camera Localization from Pixels to Pose , 2021, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[18]  Luis Merino,et al.  DLL: Direct LIDAR Localization. A map-based localization approach for aerial robots , 2021, 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[19]  V. Prisacariu,et al.  NeRF-: Neural Radiance Fields Without Known Camera Parameters , 2021, ArXiv.

[20]  Jonathan T. Barron,et al.  iNeRF: Inverting Neural Radiance Fields for Pose Estimation , 2020, 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[21]  Jonathan T. Barron,et al.  NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collections , 2020, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[22]  Kai Zhang,et al.  NeRF++: Analyzing and Improving Neural Radiance Fields , 2020, ArXiv.

[23]  Ashish Kapoor,et al.  TartanAir: A Dataset to Push the Limits of Visual SLAM , 2020, 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[24]  Pratul P. Srinivasan,et al.  NeRF , 2020, ECCV.

[25]  Tomasz Malisiewicz,et al.  SuperGlue: Learning Feature Matching With Graph Neural Networks , 2019, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[26]  Taghi M. Khoshgoftaar,et al.  A survey on Image Data Augmentation for Deep Learning , 2019, Journal of Big Data.

[27]  Torsten Sattler,et al.  Efficient & Effective Prioritized Matching for Large-Scale Image-Based Localization , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[28]  Ashish Kapoor,et al.  AirSim: High-Fidelity Visual and Physical Simulation for Autonomous Vehicles , 2017, FSR.

[29]  Jian Sun,et al.  Accelerating Very Deep Convolutional Networks for Classification and Detection , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[30]  Roberto Cipolla,et al.  PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).