Parallel Inversion of Neural Radiance Fields for Robust Pose Estimation
Not on the map. Why?
Abstract (arXiv)
We present a parallelized optimization method based on fast Neural Radiance Fields (NeRF) for estimating 6-DoF pose of a camera with respect to an object or scene. Given a single observed RGB image of the target, we can predict the translation and rotation of the camera by minimizing the residual between pixels rendered from a fast NeRF model and pixels in the observed image. We integrate a momentum-based camera extrinsic optimization procedure into Instant Neural Graphics Primitives, a recent exceptionally fast NeRF implementation. By introducing parallel Monte Carlo sampling into the pose estimation task, our method overcomes local minima and improves efficiency in a more extensive search space. We also show the importance of adopting a more robust pixel-based loss function to reduce error. Experiments demonstrate that our method can achieve improved generalization and robustness on both synthetic and real-world benchmarks.
Key facts. Cited by 55 papers in the map (data through 14 Oct 2026). Builds on 44 of its 63 references that are in the map. First posted on arXiv 18 Oct 2022.
Builds on
44 of 63 references.
References · 44
- 2014Adam: A Method for Stochastic Optimizationcited by 37,240
- 2022Instant Neural Graphics Primitives with a Multiresolution Hash Encodingcited by 3,723
- 2015Fast R-CNNcited by 2,933
- 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,689
- 2018Occupancy Networks: Learning 3D Reconstruction in Function Spacecited by 2,041
- 2021PlenOctrees for Real-time Rendering of Neural Radiance Fieldscited by 905
- 2017PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenescited by 775
- 2020Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenescited by 751
- 2021KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPscited by 673
- 2019Local Light Field Fusion: Practical View Synthesis with Prescriptive Sampling Guidelinescited by 672
- 2021BARF: Bundle-Adjusting Neural Radiance Fieldscited by 642
- 2021Neural Fields in Visual Computing and Beyondcited by 540
32 more
- 2018PVNet: Pixel-wise Voting Network for 6DoF Pose Estimationcited by 385
- 2020INeRF: Inverting Neural Radiance Fields for Pose Estimationcited by 382
- 2019DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusioncited by 355
- 2021NeRF in the Dark: High Dynamic Range View Synthesis from Noisy Raw Imagescited by 336
- 2021PhySG: Inverse Rendering with Spherical Gaussians for Physics-based Material Editing and Relightingcited by 335
- 2018DeepIM: Deep Iterative Matching for 6D Pose Estimationcited by 283
- 2018BOP: Benchmark for 6D Object Pose Estimationcited by 254
- 2018Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objectscited by 241
- 2020MetaSDF: Meta-learning Signed Distance Functionscited by 214
- 2019PVN3D: A Deep Point-wise 3D Keypoints Voting Network for 6DoF Pose Estimationcited by 204
- 2019Implicit 3D Orientation Learning for 6D Object Detection from RGB Imagescited by 170
- 2022OnePose: One-Shot Object Pose Estimation without CAD Modelscited by 157
- 2020Objectron: A Large Scale Dataset of Object-Centric Videos in the Wild with Pose Annotationscited by 155
- 2021NeRF--: Neural Radiance Fields Without Known Camera Parameterscited by 155
- 2021FFB6D: A Full Flow Bidirectional Fusion Network for 6D Pose Estimationcited by 144
- 2022Gen6D: Generalizable Model-Free 6-DoF Object Pose Estimation from RGB Imagescited by 121
- 2021Dex-NeRF: Using a Neural Radiance Field to Grasp Transparent Objectscited by 119
- 20213D Neural Scene Representations for Visuomotor Controlcited by 111
- 2021SurfEmb: Dense and Continuous Correspondence Distributions for Object Pose Estimation with Learnt Surface Embeddingscited by 95
- 2019PoseRBPF: A Rao-Blackwellized Particle Filter for 6D Object Pose Trackingcited by 91
- 2020se(3)-TrackNet: Data-driven 6D Pose Tracking by Calibrating Image Residuals in Synthetic Domainscited by 90
- 2021BundleTrack: 6D Pose Tracking for Novel Objects without Instance or Category-Level 3D Modelscited by 87
- 2022NeRF-Supervision: Learning Dense Object Descriptors from Neural Radiance Fieldscited by 86
- 2022OSOP: A Multi-Stage One Shot Object Pose Estimation Frameworkcited by 81
- 2022Loc-NeRF: Monte Carlo Localization using Neural Radiance Fieldscited by 69
- 2016Mean Absolute Percentage Error for regression modelscited by 58
- 2022NeRF-Pose: A First-Reconstruct-Then-Regress Approach for Weakly-supervised 6D Object Pose Estimationcited by 58
- 2021FiG-NeRF: Figure-Ground Neural Radiance Fields for 3D Object Category Modellingcited by 47
- 2021Single-Stage Keypoint-Based Category-Level Object Pose Estimation from an RGB Imagecited by 42
- 2020Robust, Occlusion-aware Pose Estimation for Objects Grasped by Adaptive Handscited by 26
- 20203D Object Detection and Pose Estimation of Unseen Objects in Color Images with Local Surface Embeddingscited by 22
- 2020MobilePose: Real-Time Pose Estimation for Unseen Objects with Weak Shape Supervisioncited by 16
Built on by
Citing papers per year: 2023: 15; 2024: 20; 2025: 13; 2026: 7.
55 papers.
Sort:
By area:
- 2024Scene Coordinate Reconstruction: Posing of Image Collections via Incremental Learning of a Relocalizercited by 73
- 2023CROSSFIRE: Camera Relocalization On Self-Supervised Features from an Implicit Representationcited by 45
- 2024GS-CPR: Efficient Camera Pose Refinement via 3D Gaussian Splattingcited by 31
- 2023POPE: 6-DoF Promptable Pose Estimation of Any Object, in Any Scene, with One Referencecited by 30
- 20246DGS: 6D Pose Estimation from a Single Image and a 3D Gaussian Splatting Modelcited by 28
- 2024The Unreasonable Effectiveness of Pre-Trained Features for Camera Pose Refinementcited by 21
- 2024Neural Fields in Robotics: A Surveycited by 21
- 2023NViST: In the Wild New View Synthesis from a Single Image with Transformerscited by 19
- 2023Learning to Estimate 6DoF Pose from Limited Data: A Few-Shot, Generalizable Approach using RGB Imagescited by 18
- 2024NeRFs in Robotics: A Surveycited by 16
- 2025From Sparse to Dense: Camera Relocalization with Scene-Specific Detector from Feature Gaussian Splattingcited by 14
- 2023AI-Generated Images as Data Source: The Dawn of Synthetic Eracited by 13
38 more
- 2023Cinematic Behavior Transfer via NeRF-based Differentiable Filmingcited by 12
- 2024Fed3DGS: Scalable 3D Gaussian Splatting with Federated Learningcited by 12
- 2023Diff-DOPE: Differentiable Deep Object Pose Estimationcited by 11
- 2023CF-NeRF: Camera Parameter Free Neural Radiance Fields with Incremental Learningcited by 11
- 2025Gaussian Splatting Feature Fields for Privacy-Preserving Visual Localizationcited by 10
- 2023BAA-NGP: Bundle-Adjusting Accelerated Neural Graphics Primitivescited by 8
- 2023Federated Learning for Large-Scale Scene Modeling with Neural Radiance Fieldscited by 8
- 2024GeoNLF: Geometry guided Pose-Free Neural LiDAR Fieldscited by 6
- 2023NeRFuser: Large-Scale Scene Representation by NeRF Fusioncited by 5
- 2024IFFNeRF: Initialisation Free and Fast 6DoF pose estimation from a single image and a NeRF modelcited by 5
- 2024Self-supervised Learning of Neural Implicit Feature Fields for Camera Pose Refinementcited by 5
- 2024Look Gauss, No Pose: Novel View Synthesis using Gaussian Splatting without Accurate Pose Initializationcited by 5
- 2024LoGS: Visual Localization via Gaussian Splatting with Fewer Training Imagescited by 5
- 2025Scalable and High-Quality Neural Implicit Representation for 3D Reconstructioncited by 5
- 2025LoD-Loc v2: Aerial Visual Localization over Low Level-of-Detail City Models using Explicit Silhouette Alignmentcited by 5
- 2025GSVisLoc: Generalizable Visual Localization for Gaussian Splatting Scene Representationscited by 5
- 2023MELON: NeRF with Unposed Images in SO(3)cited by 4
- 2024Marrying NeRF with Feature Matching for One-step Pose Estimationcited by 3
- 2024Intraoperative Registration by Cross-Modal Inverse Neural Renderingcited by 3
- 2025SplatPose: Geometry-Aware 6-DoF Pose Estimation from Single RGB Image via 3D Gaussian Splattingcited by 3
- 2025Can we make NeRF-based visual localization privacy-preserving?cited by 3
- 2023CaLDiff: Camera Localization in NeRF via Pose Diffusioncited by 2
- 2023Density Uncertainty Quantification with NeRF-Ensembles: Impact of Data and Scene Constraintscited by 1
- 2024ICON: Incremental CONfidence for Joint Pose and Radiance Field Optimizationcited by 1
- 2024HoloGS: Instant Depth-based 3D Gaussian Splatting with Microsoft HoloLens 2cited by 1
- 2024NeRFDeformer: NeRF Transformation from a Single View via 3D Scene Flowscited by 1
- 2025SPLATART: Articulated Gaussian Splatting with Estimated Object Structurecited by 1
- 2025Camera Pose Refinement via 3D Gaussian Splattingcited by 1
- 2026ImLoc: Revisiting Visual Localization with Image-based Representationcited by 1
- 2026Rethinking Pose Refinement in 3D Gaussian Splatting under Pose Prior and Geometric Uncertaintycited by 1
- 2023BID-NeRF: RGB-D image pose estimation with inverted Neural Radiance Fieldscited by 0
- 2024DPPE: Dense Pose Estimation in a Plenoxels Environment using Gradient Approximationcited by 0
- 2024OmniPose6D: Towards Short-Term Object Pose Tracking in Dynamic Scenes from Monocular RGBcited by 0
- 2024KeyGS: A Keyframe-Centric Gaussian Splatting Method for Monocular Image Sequencescited by 0
- 2025ImplicitCell: Resolution Cell Modeling of Joint Implicit Volume Reconstruction and Pose Refinement in Freehand 3D Ultrasoundcited by 0
- 2025GSFeatLoc: Visual Localization Using Feature Correspondence on 3D Gaussian Splattingcited by 0
- 2025Robust Visual Localization via Semantic-Guided Multi-Scale Transformercited by 0
- 2025Embodied Spatial Intelligence: from Implicit Scene Modeling to Spatial Reasoningcited by 0
50 most cited of 55 All 55 (JSON)
- 2026NeRF-based Spacecraft Reconstruction from Monocular Imagery Under Illumination Variability and Pose Uncertainty18 May 2026
- 2026PCM-NeRF: Probabilistic Camera Modeling for Neural Radiance Fields under Pose Uncertainty20 Apr 2026
- 2026Hierarchical Visual Relocalization with Nearest View Synthesis from Feature Gaussian Splatting31 Mar 2026
- 2026LoD-Loc v3: Generalized Aerial Localization in Dense Cities using Instance Silhouette Alignment20 Mar 2026
- 2026Rethinking Pose Refinement in 3D Gaussian Splatting under Pose Prior and Geometric Uncertainty17 Mar 2026
- 2026Spectral-Geometric Neural Fields for Pose-Free LiDAR View Synthesis13 Mar 2026
- 2026ImLoc: Revisiting Visual Localization with Image-based Representation7 Jan 2026
- 2025Embodied Spatial Intelligence: from Implicit Scene Modeling to Spatial Reasoning30 Aug 2025
- 2025Can we make NeRF-based visual localization privacy-preserving?26 Aug 2025
- 2025GSVisLoc: Generalizable Visual Localization for Gaussian Splatting Scene Representations25 Aug 2025
- 2025Camera Pose Refinement via 3D Gaussian Splatting25 Aug 2025
- 2025Gaussian Splatting Feature Fields for Privacy-Preserving Visual Localization31 Jul 2025
38 more
- 2025LoD-Loc v2: Aerial Visual Localization over Low Level-of-Detail City Models using Explicit Silhouette Alignment1 Jul 2025
- 2025SPLATART: Articulated Gaussian Splatting with Estimated Object Structure13 Jun 2025
- 2025Robust Visual Localization via Semantic-Guided Multi-Scale Transformer10 Jun 2025
- 2025GSFeatLoc: Visual Localization Using Feature Correspondence on 3D Gaussian Splatting29 Apr 2025
- 2025From Sparse to Dense: Camera Relocalization with Scene-Specific Detector from Feature Gaussian Splatting25 Mar 2025
- 2025ImplicitCell: Resolution Cell Modeling of Joint Implicit Volume Reconstruction and Pose Refinement in Freehand 3D Ultrasound9 Mar 2025
- 2025SplatPose: Geometry-Aware 6-DoF Pose Estimation from Single RGB Image via 3D Gaussian Splatting7 Mar 2025
- 2025Scalable and High-Quality Neural Implicit Representation for 3D Reconstruction15 Jan 2025
- 2024KeyGS: A Keyframe-Centric Gaussian Splatting Method for Monocular Image Sequences30 Dec 2024
- 2024Neural Fields in Robotics: A Survey26 Oct 2024
- 2024LoGS: Visual Localization via Gaussian Splatting with Fewer Training Images15 Oct 2024
- 2024Look Gauss, No Pose: Novel View Synthesis using Gaussian Splatting without Accurate Pose Initialization11 Oct 2024
- 2024OmniPose6D: Towards Short-Term Object Pose Tracking in Dynamic Scenes from Monocular RGB9 Oct 2024
- 2024Intraoperative Registration by Cross-Modal Inverse Neural Rendering18 Sep 2024
- 2024GS-CPR: Efficient Camera Pose Refinement via 3D Gaussian Splatting20 Aug 2024
- 20246DGS: 6D Pose Estimation from a Single Image and a 3D Gaussian Splatting Model22 Jul 2024
- 2024GeoNLF: Geometry guided Pose-Free Neural LiDAR Fields8 Jul 2024
- 2024NeRFDeformer: NeRF Transformation from a Single View via 3D Scene Flows15 Jun 2024
- 2024Self-supervised Learning of Neural Implicit Feature Fields for Camera Pose Refinement12 Jun 2024
- 2024HoloGS: Instant Depth-based 3D Gaussian Splatting with Microsoft HoloLens 23 May 2024
- 2024NeRFs in Robotics: A Survey2 May 2024
- 2024Scene Coordinate Reconstruction: Posing of Image Collections via Incremental Learning of a Relocalizer22 Apr 2024
- 2024The Unreasonable Effectiveness of Pre-Trained Features for Camera Pose Refinement16 Apr 2024
- 2024Marrying NeRF with Feature Matching for One-step Pose Estimation1 Apr 2024
- 2024IFFNeRF: Initialisation Free and Fast 6DoF pose estimation from a single image and a NeRF model19 Mar 2024
- 2024Fed3DGS: Scalable 3D Gaussian Splatting with Federated Learning18 Mar 2024
- 2024DPPE: Dense Pose Estimation in a Plenoxels Environment using Gradient Approximation16 Mar 2024
- 2024ICON: Incremental CONfidence for Joint Pose and Radiance Field Optimization17 Jan 2024
- 2023CaLDiff: Camera Localization in NeRF via Pose Diffusion23 Dec 2023
- 2023Density Uncertainty Quantification with NeRF-Ensembles: Impact of Data and Scene Constraints22 Dec 2023
- 2023CF-NeRF: Camera Parameter Free Neural Radiance Fields with Incremental Learning14 Dec 2023
- 2023NViST: In the Wild New View Synthesis from a Single Image with Transformers13 Dec 2023
- 2023Cinematic Behavior Transfer via NeRF-based Differentiable Filming29 Nov 2023
- 2023BID-NeRF: RGB-D image pose estimation with inverted Neural Radiance Fields5 Oct 2023
- 2023AI-Generated Images as Data Source: The Dawn of Synthetic Era3 Oct 2023
- 2023Diff-DOPE: Differentiable Deep Object Pose Estimation30 Sep 2023
- 2023Federated Learning for Large-Scale Scene Modeling with Neural Radiance Fields12 Sep 2023
- 2023Learning to Estimate 6DoF Pose from Limited Data: A Few-Shot, Generalizable Approach using RGB Images13 Jun 2023
50 newest of 55 All 55 (JSON)
- 2023CaLDiff: Camera Localization in NeRF via Pose Diffusioncited by 2
- 2023CaLDiff: Camera Localization in NeRF via Pose Diffusion23 Dec 2023
- 2024Scene Coordinate Reconstruction: Posing of Image Collections via Incremental Learning of a Relocalizercited by 73
- 2023CROSSFIRE: Camera Relocalization On Self-Supervised Features from an Implicit Representationcited by 45
- 2024GS-CPR: Efficient Camera Pose Refinement via 3D Gaussian Splattingcited by 31
- 2023POPE: 6-DoF Promptable Pose Estimation of Any Object, in Any Scene, with One Referencecited by 30
- 20246DGS: 6D Pose Estimation from a Single Image and a 3D Gaussian Splatting Modelcited by 28
- 2024The Unreasonable Effectiveness of Pre-Trained Features for Camera Pose Refinementcited by 21
- 2024Neural Fields in Robotics: A Surveycited by 21
- 2023NViST: In the Wild New View Synthesis from a Single Image with Transformerscited by 19
- 2023Learning to Estimate 6DoF Pose from Limited Data: A Few-Shot, Generalizable Approach using RGB Imagescited by 18
- 2024NeRFs in Robotics: A Surveycited by 16
- 2025From Sparse to Dense: Camera Relocalization with Scene-Specific Detector from Feature Gaussian Splattingcited by 14
- 2023AI-Generated Images as Data Source: The Dawn of Synthetic Eracited by 13
12 most cited of 53 All 53 (JSON)
- 2026NeRF-based Spacecraft Reconstruction from Monocular Imagery Under Illumination Variability and Pose Uncertainty18 May 2026
- 2026PCM-NeRF: Probabilistic Camera Modeling for Neural Radiance Fields under Pose Uncertainty20 Apr 2026
- 2026Hierarchical Visual Relocalization with Nearest View Synthesis from Feature Gaussian Splatting31 Mar 2026
- 2026LoD-Loc v3: Generalized Aerial Localization in Dense Cities using Instance Silhouette Alignment20 Mar 2026
- 2026Rethinking Pose Refinement in 3D Gaussian Splatting under Pose Prior and Geometric Uncertainty17 Mar 2026
- 2026Spectral-Geometric Neural Fields for Pose-Free LiDAR View Synthesis13 Mar 2026
- 2026ImLoc: Revisiting Visual Localization with Image-based Representation7 Jan 2026
- 2025Can we make NeRF-based visual localization privacy-preserving?26 Aug 2025
- 2025GSVisLoc: Generalizable Visual Localization for Gaussian Splatting Scene Representations25 Aug 2025
- 2025Camera Pose Refinement via 3D Gaussian Splatting25 Aug 2025
- 2025Gaussian Splatting Feature Fields for Privacy-Preserving Visual Localization31 Jul 2025
- 2025LoD-Loc v2: Aerial Visual Localization over Low Level-of-Detail City Models using Explicit Silhouette Alignment1 Jul 2025
12 newest of 53 All 53 (JSON)