COLMAP-Free 3D Gaussian Splatting
Not on the map. Why?
Abstract (arXiv)
While neural rendering has led to impressive advances in scene reconstruction and novel view synthesis, it relies heavily on accurately pre-computed camera poses. To relax this constraint, multiple efforts have been made to train Neural Radiance Fields (NeRFs) without pre-processed camera poses. However, the implicit representations of NeRFs provide extra challenges to optimize the 3D structure and camera poses at the same time. On the other hand, the recently proposed 3D Gaussian Splatting provides new opportunities given its explicit point cloud representations. This paper leverages both the explicit geometric representation and the continuity of the input video stream to perform novel view synthesis without any SfM preprocessing. We process the input frames in a sequential manner and progressively grow the 3D Gaussians set by taking one input frame at a time, without the need to pre-compute the camera poses. Our method significantly improves over previous approaches in view synthesis and camera pose estimation under large motion changes. Our project page is https://oasisyang.github.io/colmap-free-3dgs
Key facts. Cited by 193 papers in the map (data through 7 Oct 2026). Builds on 47 of its 60 references that are in the map. First posted on arXiv 12 Dec 2023.
Builds on
47 of 60 references.
References · 47
- 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
- 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
- 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,141
- 2022Instant Neural Graphics Primitives with a Multiresolution Hash Encodingcited by 3,698
- 2020Implicit Neural Representations with Periodic Activation Functionscited by 2,214
- 2021Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fieldscited by 2,013
- 2021Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fieldscited by 1,957
- 2021Vision Transformers for Dense Predictioncited by 1,789
- 2021Plenoxels: Radiance Fields without Neural Networkscited by 1,607
- 20234D Gaussian Splatting for Real-Time Dynamic Scene Renderingcited by 1,144
- 2015ORB-SLAM: a Versatile and Accurate Monocular SLAM Systemcited by 983
- 2021Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstructioncited by 961
35 more
- 2020Neural Sparse Voxel Fieldscited by 959
- 2021PlenOctrees for Real-time Rendering of Neural Radiance Fieldscited by 904
- 2020NeRF++: Analyzing and Improving Neural Radiance Fieldscited by 857
- 2023Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesiscited by 731
- 2021KiloNeRF: Speeding up Neural Radiance Fields with Thousands of Tiny MLPscited by 672
- 2023Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fieldscited by 667
- 2021Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstructioncited by 646
- 2023ZoeDepth: Zero-shot Transfer by Combining Relative and Metric Depthcited by 646
- 2021BARF: Bundle-Adjusting Neural Radiance Fieldscited by 641
- 2021RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from Sparse Inputscited by 614
- 2021Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance Fieldscited by 569
- 2021FastNeRF: High-Fidelity Neural Rendering at 200FPScited by 510
- 2022Point-NeRF: Point-based Neural Radiance Fieldscited by 503
- 2020INeRF: Inverting Neural Radiance Fields for Pose Estimationcited by 381
- 2023FreeNeRF: Improving Few-shot Neural Rendering with Free Frequency Regularizationcited by 310
- 2020Single-View View Synthesis with Multiplane Imagescited by 263
- 2019Differentiable Surface Splatting for Point-based Geometry Processingcited by 238
- 2022NoPe-NeRF: Optimising Neural Radiance Field with No Pose Priorcited by 235
- 2020Free View Synthesiscited by 231
- 2021Self-Calibrating Neural Radiance Fieldscited by 204
- 2021InfoNeRF: Ray Entropy Minimization for Few-Shot Neural Volume Renderingcited by 180
- 2023HyperReel: High-Fidelity 6-DoF Video with Ray-Conditioned Samplingcited by 173
- 2020Stable View Synthesiscited by 165
- 2022SinNeRF: Training Neural Radiance Fields on Complex Scenes from a Single Imagecited by 165
- 2021MINE: Towards Continuous Depth MPI with NeRF for Novel View Synthesiscited by 133
- 2023Progressively Optimized Local Radiance Fields for Robust View Synthesiscited by 103
- 2022Differentiable Point-Based Radiance Fields for Efficient View Synthesiscited by 86
- 2023Neural Point Catacaustics for Novel-View Synthesis of Reflectionscited by 81
- 2020Worldsheet: Wrapping the World in a 3D Sheet for View Synthesis from a Single Imagecited by 77
- 2023NeO 360: Neural Fields for Sparse View Synthesis of Outdoor Scenescited by 53
- 2022SiNeRF: Sinusoidal Neural Radiance Fields for Joint Pose Estimation and Scene Reconstructioncited by 49
- 2022RUST: Latent Neural Scene Representations from Unposed Imagerycited by 41
- 2023FlowCam: Training Generalizable 3D Radiance Fields without Camera Poses via Pixel-Aligned Scene Flowcited by 38
- 2023LU-NeRF: Scene and Pose Estimation by Synchronizing Local Unposed NeRFscited by 33
- 2022MonoNeRF: Learning Generalizable NeRFs from Monocular Videos without Camera Posecited by 10
Built on by
Citing papers per year: 2024: 64; 2025: 94; 2026: 35.
193 papers.
Sort:
By area:
- 2024No Pose, No Problem: Surprisingly Simple 3D Gaussian Splats from Sparse Unposed Imagescited by 241
- 2025AnySplat: Feed-forward 3D Gaussian Splatting from Unconstrained Viewscited by 234
- 2024MegaSaM: Accurate, Fast, and Robust Structure and Motion from Casual Dynamic Videoscited by 224
- 2024A Survey on 3D Gaussian Splattingcited by 221
- 2025FLARE: Feed-forward Geometry, Appearance and Camera Estimation from Uncalibrated Sparse Viewscited by 176
- 2024Recent Advances in 3D Gaussian Splattingcited by 130
- 2024ManiGaussian: Dynamic Gaussian Splatting for Multi-task Robotic Manipulationcited by 106
- 2024Stereo4D: Learning How Things Move in 3D from Internet Stereo Videoscited by 89
- 20243D Gaussian Splatting: Survey, Technologies, Challenges, and Opportunitiescited by 77
- 2025RayZer: A Self-supervised Large View Synthesis Modelcited by 74
- 2024InstantSplat: Sparse-view Gaussian Splatting in Secondscited by 67
- 2024GS2Mesh: Surface Reconstruction from Gaussian Splatting via Novel Stereo Viewscited by 65
38 more
- 2024MotionGS: Exploring Explicit Motion Guidance for Deformable 3D Gaussian Splattingcited by 62
- 2024BAD-Gaussians: Bundle Adjusted Deblur Gaussian Splattingcited by 61
- 2024FlowMap: High-Quality Camera Poses, Intrinsics, and Depth via Gradient Descentcited by 61
- 2024PF3plat: Pose-Free Feed-Forward 3D Gaussian Splattingcited by 60
- 2024GIC: Gaussian-Informed Continuum for Physical Property Identification and Simulationcited by 48
- 2024GFlow: Recovering 4D World from Monocular Videocited by 46
- 2024EG4D: Explicit Generation of 4D Object without Score Distillationcited by 45
- 2024SplineGS: Robust Motion-Adaptive Spline for Real-Time Dynamic 3D Gaussians from Monocular Videocited by 44
- 2024GraspSplats: Efficient Manipulation with 3D Feature Splattingcited by 41
- 2024Gaussian Splatting: 3D Reconstruction and Novel View Synthesis, a Reviewcited by 40
- 2024SelfSplat: Pose-Free and 3D Prior-Free Generalizable 3D Gaussian Splattingcited by 40
- 2025LangSplatV2: High-dimensional 3D Language Gaussian Splatting with 450+ FPScited by 40
- 2025No Pose at All: Self-Supervised Pose-Free 3D Gaussian Splatting from Sparse Viewscited by 38
- 2025GWM: Towards Scalable Gaussian World Models for Robotic Manipulationcited by 37
- 2025Instant Gaussian Stream: Fast and Generalizable Streaming of Dynamic Scene Reconstruction via Gaussian Splattingcited by 35
- 2024GaussianHeads: End-to-End Learning of Drivable Gaussian Head Avatars from Coarse-to-fine Representationscited by 31
- 2024Gaussian Splatting on the Move: Blur and Rolling Shutter Compensation for Natural Camera Motioncited by 29
- 2025What Is The Best 3D Scene Representation for Robotics? From Geometric to Foundation Modelscited by 29
- 2025FlowR: Flowing from Sparse to Dense 3D Reconstructionscited by 28
- 2024Sort-free Gaussian Splatting via Weighted Sum Renderingcited by 27
- 2025Luminance-GS: Adapting 3D Gaussian Splatting to Challenging Lighting Conditions with View-Adaptive Curve Adjustmentcited by 27
- 2024Feat2GS: Probing Visual Foundation Models with Gaussian Splattingcited by 26
- 2024GauStudio: A Modular Framework for 3D Gaussian Splatting and Beyondcited by 24
- 2024BillBoard Splatting (BBSplat): Learnable Textured Primitives for Novel View Synthesiscited by 23
- 2025LongSplat: Robust Unposed 3D Gaussian Splatting for Casual Long Videoscited by 23
- 2024MM3DGS SLAM: Multi-modal 3D Gaussian Splatting for SLAM Using Vision, Depth, and Inertial Measurementscited by 22
- 2024Event3DGS: Event-Based 3D Gaussian Splatting for High-Speed Robot Egomotioncited by 22
- 2025Outdoor Monocular SLAM with Global Scale-Consistent 3D Gaussian Pointmapscited by 21
- 2024GGRt: Towards Pose-free Generalizable 3D Gaussian Splatting in Real-timecited by 20
- 2024GaussianSR: 3D Gaussian Super-Resolution with 2D Diffusion Priorscited by 20
- 2024Free-SurGS: SfM-Free 3D Gaussian Splatting for Surgical Scene Reconstructioncited by 20
- 2024IncEventGS: Pose-Free Gaussian Splatting from a Single Event Cameracited by 20
- 2024Touch-GS: Visual-Tactile Supervised 3D Gaussian Splattingcited by 16
- 2024VDG: Vision-Only Dynamic Gaussian for Driving Simulationcited by 16
- 2024HDRSplat: Gaussian Splatting for High Dynamic Range 3D Scene Reconstruction from Raw Imagescited by 15
- 2024EF-3DGS: Event-Aided Free-Trajectory 3D Gaussian Splattingcited by 15
- 2024Gaussian Splatting for Efficient Satellite Image Photogrammetrycited by 15
- 2025VicaSplat: A Single Run is All You Need for 3D Gaussian Splatting and Camera Estimation from Unposed Video Framescited by 15
50 most cited of 193 All 193 (JSON)
- 2026VoxelTTO: Voxel-Aligned Feed-Forward 3D Gaussian Splatting with Test-Time Optimization18 Sep 2026
- 2026CADSplat: Sparse-View 3D Gaussian Splatting Aided by CAD Models for Robust, Photorealistic Digital-Twin Reconstruction16 Sep 2026
- 2026Triangulation-Free Bundle Adjustment with Graduated Non-Convexity for Camera Pose Refinement from Coarse Priors21 Aug 2026
- 2026Mirror Learning30 Jul 2026
- 20264DHumanDiff: Direct Text-to-4DGS Generation for Consistent 360-Degree Dynamic Humans30 Jul 2026
- 20263DGBGS: 3D Granular Ball Gaussian Splatting for Compact Novel View Synthesis29 Jul 2026
- 2026Event3R: Asynchronous-to-Global 3D Reconstruction from Event Camera via Spatial-Temporal Feature Aggregation17 Jul 2026
- 2026AsySplat: Efficient Asymmetric 3D Gaussian Splatting for Long-Sequence Scene Modeling13 Jul 2026
- 2026Dynamic Inverse Rendering for Enhanced Material-Lighting Decomposition10 Jul 2026
- 2026NoDrift3R: Raymap-Guided Coupling for Drift-Robust Unposed Feed-Forward 3D Reconstruction8 Jul 2026
- 2026UniSpine-GS: An Efficient Physics-Aware Gaussian Framework for Cross-Modality Multi-view Spine Image Synthesis6 Jul 2026
- 2026StructSplat: Generalizable 3D Gaussian Splatting from Uncalibrated Sparse Views26 Jun 2026
38 more
- 2026Learning Stable Canonical Worlds for Novel View Synthesis and Beyond22 Jun 2026
- 2026Point-Cloud-Assistant Localized Statistical Channel Prediction by Tangent Gaussian Splatting17 Jun 2026
- 2026ZipSplat: Fewer Gaussians, Better Splats3 Jun 2026
- 2026A Cookbook of 3D Vision: Data, Learning Paradigms, and Application2 Jun 2026
- 2026RayDer: Scalable Self-Supervised Novel View Synthesis from Real-World Video29 May 2026
- 2026ForeSplat: Optimization-Aware Foresight for Feed-Forward 3D Gaussian Splatting21 May 2026
- 2026Towards Physically Consistent 4D Scene Reconstruction for Closed-loop Autonomous Driving Simulation20 May 2026
- 20263DReflecNet: A Large-Scale Dataset for 3D Reconstruction of Reflective, Transparent, and Low-Texture Objects11 May 2026
- 2026Kinematics-Driven Gaussian Shape Deformation for Blurry Monocular Dynamic Scenes9 May 2026
- 2026Learning 3D Representations for Spatial Intelligence from Unposed Multi-View Images12 Apr 2026
- 2026GLINT: Modeling Scene-Scale Transparency via Gaussian Radiance Transport27 Mar 2026
- 2026AirSplat: Alignment and Rating for Robust Feed-Forward 3D Gaussian Splatting26 Mar 2026
- 2026Pose-Free Omnidirectional Gaussian Splatting for 360-Degree Videos with Consistent Depth Priors24 Mar 2026
- 2026GeoNVS: Geometry Grounded Video Diffusion for Novel View Synthesis16 Mar 2026
- 2026Dark3R: Learning Structure from Motion in the Dark5 Mar 2026
- 2026GloSplat: Joint Pose-Appearance Optimization for Faster and More Accurate 3D Reconstruction5 Mar 2026
- 2026Intrinsic Geometry-Appearance Consistency Optimization for Sparse-View Gaussian Splatting3 Mar 2026
- 2026Pseudo-View Enhancement via Confidence Fusion for Unposed Sparse-View Reconstruction25 Feb 2026
- 2026Event-Aided Sharp Radiance Field Reconstruction for Fast-Flying Drones24 Feb 2026
- 2026Dropping Anchor and Spherical Harmonics for Sparse-view Gaussian Splatting24 Feb 2026
- 2026Unifying Color and Lightness Correction with View-Adaptive Curve Adjustment for Robust 3D Novel View Synthesis20 Feb 2026
- 2026CompSplat: Compression-aware 3D Gaussian Splatting for Real-world Video10 Feb 2026
- 2026Pi-GS: Sparse-View Gaussian Splatting with Dense π3 Initialization3 Feb 2026
- 2025Improved 3D Gaussian Splatting of Unknown Spacecraft Structure Using Space Environment Illumination Knowledge30 Dec 2025
- 2025G3Splat: Geometrically Consistent Generalizable Gaussian Splatting19 Dec 2025
- 2025Off The Grid: Detection of Primitives for Feed-Forward 3D Gaussian Splatting17 Dec 2025
- 2025Splannequin: Freezing Monocular Mannequin-Challenge Footage with Dual-Detection Splatting4 Dec 2025
- 2025What Is The Best 3D Scene Representation for Robotics? From Geometric to Foundation Models3 Dec 2025
- 2025SurfFill: Completion of LiDAR Point Clouds via Gaussian Surfel Splatting2 Dec 2025
- 2025TALO: Pushing 3D Vision Foundation Models Towards Globally Consistent Online Reconstruction2 Dec 2025
- 2025FACT-GS: Frequency-Aligned Complexity-Aware Texture Reparameterization for 2D Gaussian Splatting28 Nov 2025
- 2025NoPe-NeRF++: Local-to-Global Optimization of NeRF with No Pose Prior21 Nov 2025
- 2025Rad-GS: Radar-Vision Integration for 3D Gaussian Splatting SLAM in Outdoor Environments20 Nov 2025
- 2025Dental3R: Geometry-Aware Pairing for Intraoral 3D Reconstruction from Sparse-View Photographs18 Nov 2025
- 2025High Dynamic Range 3D Gaussian Splatting via Luminance-Chromaticity Decomposition17 Nov 2025
- 20254D3R: Motion-Aware Neural Reconstruction and Rendering of Dynamic Scenes from Monocular Videos7 Nov 2025
- 2025DentalSplat: Dental Occlusion Novel View Synthesis from Sparse Intra-Oral Photographs5 Nov 2025
- 2025EA3D: Online Open-World 3D Object Extraction from Streaming Videos29 Oct 2025
50 newest of 193 All 193 (JSON)
- 2024EG4D: Explicit Generation of 4D Object without Score Distillationcited by 45
- 2024GaussianSR: 3D Gaussian Super-Resolution with 2D Diffusion Priorscited by 20
- 2024Gaussian Scenes: Pose-Free Sparse-View Scene Reconstruction using Depth-Enhanced Diffusion Priorscited by 8
- 2026GeoNVS: Geometry Grounded Video Diffusion for Novel View Synthesiscited by 3
- 2026Intrinsic Geometry-Appearance Consistency Optimization for Sparse-View Gaussian Splattingcited by 2
- 2025AxisPose: Model-Free Matching-Free Single-Shot 6D Object Pose Estimation via Axis Generationcited by 0
- 2025OGGSplat: Open Gaussian Growing for Generalizable Reconstruction with Expanded Field-of-Viewcited by 0
- 2026GeoNVS: Geometry Grounded Video Diffusion for Novel View Synthesis16 Mar 2026
- 2026Intrinsic Geometry-Appearance Consistency Optimization for Sparse-View Gaussian Splatting3 Mar 2026
- 2025OGGSplat: Open Gaussian Growing for Generalizable Reconstruction with Expanded Field-of-View5 Jun 2025
- 2025AxisPose: Model-Free Matching-Free Single-Shot 6D Object Pose Estimation via Axis Generation9 Mar 2025
- 2024Gaussian Scenes: Pose-Free Sparse-View Scene Reconstruction using Depth-Enhanced Diffusion Priors24 Nov 2024
- 2024GaussianSR: 3D Gaussian Super-Resolution with 2D Diffusion Priors14 Jun 2024
- 2024EG4D: Explicit Generation of 4D Object without Score Distillation28 May 2024
- 2026Mirror Learningcited by 0
- 2026Mirror Learning30 Jul 2026
- 2024No Pose, No Problem: Surprisingly Simple 3D Gaussian Splats from Sparse Unposed Imagescited by 241
- 2025AnySplat: Feed-forward 3D Gaussian Splatting from Unconstrained Viewscited by 234
- 2024MegaSaM: Accurate, Fast, and Robust Structure and Motion from Casual Dynamic Videoscited by 224
- 2024A Survey on 3D Gaussian Splattingcited by 221
- 2025FLARE: Feed-forward Geometry, Appearance and Camera Estimation from Uncalibrated Sparse Viewscited by 176
- 2024Recent Advances in 3D Gaussian Splattingcited by 130
- 2024ManiGaussian: Dynamic Gaussian Splatting for Multi-task Robotic Manipulationcited by 106
- 2024Stereo4D: Learning How Things Move in 3D from Internet Stereo Videoscited by 89
- 20243D Gaussian Splatting: Survey, Technologies, Challenges, and Opportunitiescited by 77
- 2025RayZer: A Self-supervised Large View Synthesis Modelcited by 74
- 2024InstantSplat: Sparse-view Gaussian Splatting in Secondscited by 67
- 2024GS2Mesh: Surface Reconstruction from Gaussian Splatting via Novel Stereo Viewscited by 65
12 most cited of 180 All 180 (JSON)
- 2026VoxelTTO: Voxel-Aligned Feed-Forward 3D Gaussian Splatting with Test-Time Optimization18 Sep 2026
- 2026CADSplat: Sparse-View 3D Gaussian Splatting Aided by CAD Models for Robust, Photorealistic Digital-Twin Reconstruction16 Sep 2026
- 2026Triangulation-Free Bundle Adjustment with Graduated Non-Convexity for Camera Pose Refinement from Coarse Priors21 Aug 2026
- 20264DHumanDiff: Direct Text-to-4DGS Generation for Consistent 360-Degree Dynamic Humans30 Jul 2026
- 20263DGBGS: 3D Granular Ball Gaussian Splatting for Compact Novel View Synthesis29 Jul 2026
- 2026Event3R: Asynchronous-to-Global 3D Reconstruction from Event Camera via Spatial-Temporal Feature Aggregation17 Jul 2026
- 2026AsySplat: Efficient Asymmetric 3D Gaussian Splatting for Long-Sequence Scene Modeling13 Jul 2026
- 2026Dynamic Inverse Rendering for Enhanced Material-Lighting Decomposition10 Jul 2026
- 2026UniSpine-GS: An Efficient Physics-Aware Gaussian Framework for Cross-Modality Multi-view Spine Image Synthesis6 Jul 2026
- 2026StructSplat: Generalizable 3D Gaussian Splatting from Uncalibrated Sparse Views26 Jun 2026
- 2026Learning Stable Canonical Worlds for Novel View Synthesis and Beyond22 Jun 2026
- 2026Point-Cloud-Assistant Localized Statistical Channel Prediction by Tangent Gaussian Splatting17 Jun 2026
12 newest of 180 All 180 (JSON)