3D Geometric Foundation Models (3R) with SLAM

Survey Paper

• Advances in Feed-Forward 3D Reconstruction and View Synthesis: A Survey, arXiv 2025. [Paper] [Website]
• Feed-Forward 3D Scene Modeling: A Problem-Driven Perspective, arXiv 2026. [Paper] [Website]

3D Geometric Foundation Models (3R)

• DUSt3R: Geometric 3D Vision Made Easy, CVPR 2024. [Paper] [Code] [Website]
• Monst3r: A simple approach for estimating geometry in the presence of motion, ICLR 2025. [Paper] [Code] [Website]
• LoRA3D: Low-Rank Self-Calibration of 3D Geometric Foundation Models, ICLR 2025. [Paper] [Website]
• (CUT3R) Continuous 3D Perception Model with Persistent State, CVPR 2025. [Paper] [Code] [Website]
• Reloc3r: Large-Scale Training of Relative Camera Pose Regression for Generalizable, Fast, and Accurate Visual Localization, CVPR 2025. [Paper] [Code]
• DAS3R: Dynamics-Aware Gaussian Splatting for Static Scene Reconstruction, arXiv 2024. [Paper] [Code] [Website]
• MASt3R-SfM: a Fully-Integrated Solution for Unconstrained Structure-from-Motion, 3DV 2025. [Paper] [Code]
• Splatt3R: Zero-shot Gaussian Splatting from Uncalibrated Image Pairs, arXiv 2024. [Paper] [Project] [Code]
• SAB3R: Semantic-Augmented Backbone in 3D Reconstruction, arXiv 2024. [Paper]
• No Pose, No Problem: Surprisingly Simple 3D Gaussian Splats from Sparse Unposed Images, ICLR 2025. [Paper] [Code] [Website]
• Doppelgangers++: Improved Visual Disambiguation with Geometric 3D Features, CVPR 2025. [Paper] [Code] [Project]
• (Spann3R) 3D Reconstruction with Spatial Memory, 3DV 2025. [Paper] [Code] [Website]
• Fast3R: Towards 3D Reconstruction of 1000+ Images in One Forward Pass, CVPR 2025. [Paper] [Website] [Code]
• InstantSplat: Unbounded Sparse-view Pose-free Gaussian Splatting in 40 Seconds, arXiv 2025. [Paper] [Website] [Code]
• SPARS3R: Semantic Prior Alignment and Regularization for Sparse 3D Reconstruction, arXiv 2024. [Paper] [Code]
• Align3R: Aligned Monocular Depth Estimation for Dynamic Videos, CVPR 2025. [Paper] [Code] [Website]
• (MASt3R) Grounding Image Matching in 3D with MASt3R, ECCV 2024. [Paper] [Code] [Website]
• VGGT: Visual Geometry Grounded Transformer, CVPR 2025. [Paper] [Code] [Website]
• E3D-Bench: A Benchmark for End-to-End 3D Geometric Foundation Models, arXiv 2025. [Paper] [Code] [Website]
• π^3: Scalable Permutation-Equivariant Visual Geometry Learning, arXiv 2025. [Paper] [Code] [Website]
• Dens3R: A Foundation Model for 3D Geometry Prediction, ICCV 2025. [Paper] [Code] [Website]
• LONG3R: Long Sequence Streaming 3D Reconstruction, ICCV 2025. [Paper] [Code] [Website]
• PanoSplatt3R: Leveraging Perspective Pretraining for Generalized Unposed Wide-Baseline Panorama Reconstruction, ICCV 2025. [Paper] [Code] [Website]
• Ov3R: Open-Vocabulary Semantic 3D Reconstruction from RGB Videos, CVPR 2026. [Paper]
• G-CUT3R: Guided 3D Reconstruction with Camera and Depth Prior Integration, arXiv 2025. [Paper]
• ViPE: Video Pose Engine for 3D Geometric Perception, arXiv 2025. [Paper] [Code] [Website]
• FastVGGT: Training-Free Acceleration of Visual Geometry Transformer, arXiv 2025. [Paper]
• SAIL-Recon: Large SfM by Augmenting Scene Regression with Localization, arXiv 2025. [Paper] [Code] [Website]
• Streaming 4D Visual Geometry Transformer, arXiv 2025. [Paper] [Code] [Website]
• MapAnything: Universal Feed-Forward Metric 3D Reconstruction, arXiv 2025. [Paper] [Code] [Website]
• TTT3R: 3D Reconstruction as Test-Time Training, arXiv 2025. [Paper] [Code] [Website]
• Co-Me: Confidence Guided Token Merging for Visual Geometric Transformers, arXiv 2025. [Paper] [Code] [Website]
• MB3R: Accurate Feed-forward Metric-scale 3D Reconstruction with Backend, arXiv 2025. [Paper]
• VG3T: Visual Geometry Grounded Gaussian Transformer, arXiv 2025. [Paper]
• KV-Tracker: Real-Time Pose Tracking with Transformers, arXiv 2025. [Paper] [Website]
• V-DPM: 4D Video Reconstruction with Dynamic Point Mapss, arXiv 2026. [Paper] [Website] [Code]
• S-MUSt3R: Sliding Multi-view 3D Reconstruction, arXiv 2026. [Paper]
• Flow3r: Factored Flow Prediction for Scalable Visual Geometry Learning, arXiv 2026. [Paper] [Website] [Code]
• VGG-T3: Offline Feed-Forward 3D Reconstruction at Scale, CVPR 2026. [Paper]
• OnlineX: Unified Online 3D Reconstruction and Understanding with Active-to-Stable State Evolution, CVPR Finding 2026 2026. [Paper] [Website]
• ZipMap: Linear-Time Stateful 3D Reconstruction with Test-Time Training, CVPR 2026. [Paper] [Code] [Website]
• Dark3R: Learning Structure from Motion in the Dark, CVPR 2026. [Paper] [Website]
• GGPT: Geometry-Grounded Point Transformer, CVPR 2026. [Paper] [Code] [Website]
• Repurposing Geometric Foundation Models for Multi-view Diffusion, arXiv 2026. [Paper] [Code] [Website]
• NOVA3R: Non-pixel-aligned Visual Transformer for Amodal 3D Reconstruction, ICLR 2026. [Paper] [Website
• StereoVGGT: A Training-Free Visual Geometry Transformer for Stereo Vision, arXiv 2026. [Paper] [Code]
• Fisheye3R: Adapting Unified 3D Feed-Forward Foundation Models to Fisheye Lenses, arXiv 2026. [Paper]
• Learning 3D Reconstruction with Priors in Test Time, arXiv 2026. [Paper] [Website
• Scal3R: Scalable Test-Time Training for Large-Scale 3D Reconstruction, CVPR 2026. [Paper] [Website [Code]
• AnyRecon: Arbitrary-View 3D Reconstruction with Video Diffusion Model, arXiv 2026. [Paper] [Website [Code]
• Emergent Extreme-View Geometry in 3D Foundation Models, CVPR 2026. [Paper] [Website [Code]
• VGGT-Ω, CVPR 2026. [Paper] [Website [Code]
• Mamba-VGGT: Persistent Long-Sequence Video Geometry Grounded Transformer via External Sliding Window Mamba Memory, arXiv 2026. [Paper]
• UNIT: Unified Geometry Learning with Group Autoregressive Transformer, arXiv 2026. [Paper] [Website [Code]
• Global Structure-from-Motion Meets Feedforward Reconstruction, CVPR 2026. [Paper] [Website [Code]
• TriSplat: Simulation-Ready Feed-Forward 3D Scene Reconstruction, arXiv 2026. [Paper] [Website] [Code]
• R3: 3D Reconstruction via Relative Regression, arXiv 2026. [Paper] [Website [Code]
• EPO: Boosting 3D Foundation Models with Edge-based Pose Optimization, ECCV 2026. [Paper]
• RayTun3R: Online Camera Adaptation in 3D Foundation Models, arXiv 2026. [Paper]
• VI3: Grounding Pretrained 3D Foundation Models with Inertial Cues, arXiv 2026. [Paper] [Website
• Hierarchical Structure-from-Motion Scales Feedforward Reconstruction, arXiv 2026. [Paper]
• RoMa-Ω: What Feed-Forward 3D Models Know About Image Matching, ECCVw 2026. [Paper] [Code]
• What VGGT Knows About Overlap: Probing Geometric Foundation Models for Co-Visibility, arXiv 2026. [Paper] [Code]
• A Calibration Audit of Confidence in Feed-Forward 3D Reconstruction Models, arXiv 2026. [Paper]

SLAM

• Hier-SLAM++: Neuro-symbolic semantic slam with a hierarchically categorical Gaussian splatting, arXiv 2025. [Paper]
• MegaSaM: Accurate, Fast, and Robust Structure and Motion from Casual Dynamic Videos, CVPR 2025. [Paper] [Website] [Code]
• SLAM3R: Real-Time Dense Scene Reconstruction from Monocular RGB Videos, CVPR 2025. [Paper] [Code]
• MASt3R-SLAM: Real-Time Dense SLAM with 3D Reconstruction Priors, CVPR 2025. [Paper] [Website] [Code]
• VGGT-SLAM: Dense RGB SLAM Optimized on the SL(4) Manifold, arXiv 2025. [Paper] [Code]
• Outdoor Monocular SLAM with Global Scale-Consistent 3D Gaussian Pointmaps, ICCV 2025. [Paper] [Website] [Code]
• VGGT-Long: Chunk it, Loop it, Align it – Pushing VGGT’s Limits on Kilometer-scale Long RGB Sequences, ICRA, 2025. [Paper] [Code]
• Pseudo Depth Meets Gaussian: A Feed-forward RGB SLAM Baseline, IROS, 2025. [Paper] [Code]
• 3D Foundation Model-Based Loop Closing for Decentralized Collaborative SLAM, RAL, 2025. [Paper]
• ViSTA-SLAM: Visual SLAM with Symmetric Two-view Association, 3DV, 2026. [Paper] [Code]
• SLAM-Former: Putting SLAM into One Transformer, arXiv 2025. [Paper] [Website] [Code]
• MASt3R-Fusion: Integrating Feed-Forward Visual Model with IMU, GNSS for High-Functionality SLAM, arXiv, 2025. [Paper] [Code]
• GRS-SLAM3R: Real-Time Dense SLAM with Gated Recurrent State, ICRA, 2026. [Paper]
• EC3R-SLAM: Efficient and Consistent Monocular Dense SLAM with Feed-Forward 3D Reconstruction, ICRA, 2026. [Paper] [Website] [Code]
• Visual Odometry with Transformers, arXiv 2025. [Paper] [Website] [Code]
• ARTDECO: Towards Efficient and High-Fidelity On-the-Fly 3D Reconstruction with Structured Scene Representation, arXiv, 2025. [Paper] [Website] [Code]
• MASt3R-GS: Bridging 3D Reconstruction Priors with Gaussian Splatting for Real-Time Dense SLAM, IROSw, 2025. [Paper]
• LiDAR-VGGT: Cross-Modal Coarse-to-Fine Fusion for Globally Consistent and Metric-Scale Dense Mapping, RAL, 2026. [Paper] [Code]
• Building temporally coherent 3D maps with VGGT for memory-efficient Semantic SLAM, arXiv, 2025. [Paper]
• SING3R-SLAM: Submap-based Indoor Monocular Gaussian SLAM with 3D Reconstruction Prior, arXiv, 2025. [Paper]
• KM-ViPE: Online Tightly Coupled Vision-Language-Geometry Fusion for Open-Vocabulary Semantic SLAM, arXiv, 2025. [Paper] [Code]
• Dynamic Visual SLAM using a General 3D Prior, CVPR, 2026. [Paper] [Code]
• OpenMonoGS-SLAM: Monocular Gaussian Splatting SLAM with Open-set Semantics, arXiv, 2025. [Paper]
• Keyframe-Based Feed-Forward Visual Odometry, arXiv 2026. [Paper]
• VGGT-SLAM 2.0: Real time Dense Feed-forward Scene Reconstruction, RSS 2026. [Paper]
• VGGT-Motion: Motion-Aware Calibration-Free Monocular SLAM for Long-Range Consistency, arXiv 2026. [Paper]
• VGGT-based online 3D semantic SLAM for indoor scene understanding and navigation, arXiv 2026. [Paper]
• VGGT-Geo: Probabilistic Geometric Fusion of Visual Geometry Grounded Transformer Priors for Robust Dense Indoor SLAM, ISPRS International Journal of Geo-Information 2026. [Paper]
• IRIS-SLAM: Unified Geo-Instance Representations for Robust Semantic Localization and Mapping, arXiv 2026. [Paper]
• AIM-SLAM: Dense Monocular SLAM via Adaptive and Informative Multi-View Keyframe Prioritization with Foundation Model, ICRA, 2026. [Paper] [Website]
• Flash-Mono: Feed-Forward Accelerated Gaussian Splatting Monocular SLAM, ICLR, 2026. [Paper] [Website]
• HyVGGT-VO: Tightly Coupled Hybrid Dense Visual Odometry with Feed-Forward Models, RAL, 2026. [Paper] [Website] [Code]
• VGGT-SLAM++, arXiv, 2026. [Paper]
• Online3R: Online Learning for Consistent Sequential Reconstruction Based on Geometry Foundation Model, CVPR, 2026. [Paper] [Website]
• Accelerating Transformer-Based Monocular SLAM via Geometric Utility Scoring, arXiv, 2026. [Paper] [Website]
• Keep It CALM: Toward Calibration-Free Kilometer-Level SLAM with Visual Geometry Foundation Models via an Assistant Eye, arXiv, 2026. [Paper] [Code]
• MR.ScaleMaster: Scale-Consistent Collaborative Mapping from Crowd-Sourced Monocular Videos, arXiv, 2026. [Paper] [Code]
• RADIO-ViPE: Online Tightly Coupled Multi-Modal Fusion for Open-Vocabulary Semantic SLAM in Dynamic Environments, arXiv, 2026. [Paper] [Website] [Code]
• MASTD3R-SLAM: Monocular Adaptive Semantic Tracking and Dynamic Reconstruction SLAM, ICRA, 2026. Paper-todo
• AMB3R: Accurate Feed-forward Metric-scale 3D Reconstruction with Backend, CVPR, 2026. [Paper] [Website] [Code]
• TALO: Pushing 3D Vision Foundation Models Towards Globally Consistent Online Reconstruction, CVPR, 2026. [Paper] [Website] [Code]
• (LingBot-Map) Geometric Context Transformer for Streaming 3D Reconstruction, arXiv 2026. [Paper] [Website [Code]
• TopoMA: Topology-Guided Multi-Agent Dense RGB 3D Reconstruction via Distributed Inference, CVPR, 2026. Paper-todo
• LEXI-SG: Monocular 3D Scene Graph Mapping with Room-Guided Feed-Forward Reconstruction, arXiv, 2026. [Paper]
• M^3: Dense Matching Meets Multi-View Foundation Models for Monocular Gaussian Splatting SLAM, arXiv, 2026. [Paper] [Website] [Code]
• Efficient Feature-Free Initialization for Monocular Visual-Inertial Systems Using a Feed-Forward 3D Model, RSS, 2026. [Paper] [Code]
• PRISM-SLAM: Probabilistic Ray-Grounded Inference for Scale-aware Metric SLAM, arXiv, 2026. [Paper] [Website]
• CoMo3R-SLAM: Collaborative Monocular Dense SLAM with Learned 3D Reconstruction Priors for Outdoor Multi-Agent Systems, arXiv, 2026. [Paper] [Website]
• ScaRF-SLAM: Scale-Consistent Reconstruction with Feed-Forward Models and Classical Visual SLAM, arXiv, 2026. [Paper] [Code]
• Anchor3R: Streaming 3D Reconstruction with Transient Anchors for Long-Horizon Visual Mapping, arXiv, 2026. [Paper]
• GeoGS-SLAM: Online Monocular Reconstruction Using Gaussian Splatting with Geometric Priors, ICRA, 2026. [Paper] [Website]
• MAGiSt3R: Multi-Agent Feed-forward 3D Reconstruction from Monocular RGB Videos, ECCV 2026. [Paper] [Website]
• Look Up and Look Back: Hidden Attention and Latent Orientation in a Frozen Foundation Model for Panoramic SLAM, arXiv, 2026. [Paper]
• UniSim-SLAM: Feed-Forward SLAM with Unified Sim(3) Optimization, ECCV, 2026. [Paper] [Website]
• SLAMFormer-∞: Infinite SLAM Transformer for Unbounded Frontend and Backend Processing, arXiv, 2026. [Paper] [Website]
• VGGT-Align: Bridging Local Reconstruction and Global Consistency for Long-Sequence 3D Reconstruction, ACM MM, 2026. [Paper] [Code]
• On-the-Fly3R: Towards Robust Online 3D Reconstruction with Feed-Forward 3R Models for Large-Scale UAV Scenarios, arXiv, 2026. [Paper] [Code]
• VOIM: Training-Free Open-Vocabulary 3D Instance Mapping for RGB-D and Monocular SLAM, arXiv, 2026. [Paper]
• RoSe-SLAM: Robust Semantic-Aware Gaussian Splatting SLAM from Dynamic Monocular Videos, IROS, 2026. [Paper]
• FoundationSLAM: Unleashing the Power of Depth Foundation Models for End-to-End Dense Visual SLAM, AAAI, 2026. [Paper]
• Functional-SLAM: Interaction-Aware Mapping with Online Functional Scene Graphs, CoRL, 2026. [Paper] [Code]
• FFVO: A Feedforward Pose Decoder for Long-Horizon Visual Odometry, arXiv, 2026. [Paper]
• SURE-Map: Self-Correcting Streaming Geometric Foundation Models, arXiv, 2026. [Paper] [Website] [Code]
• VGGT-GS SLAM: Uncalibrated Monocular Gaussian Splatting SLAM with Feed-Forward Priors, arXiv, 2026. [Paper]
• AMB3R-SLAM: Kilometer-scale SLAM with Hierarchical Backend, arXiv, 2026. [Paper] [Website] [Code]
• GRF-Recon: Global Ray-Field Optimization for Long-Sequence Feed-forward Reconstruction, arXiv, 2026. [Paper]
• Dense Monocular SLAM in Real-Time With Structured Gaussian Representation, RAL, 2026. [Paper]
• Adaptive World Memory 3D Foundation Model for Scalable 3D Mapping, Localization, and Rendering, arXiv, 2026. [Paper] [Code]
• NOCTIF3R: Feed-Forward Monocular Real-Time SLAM for Photon-Limited Scenes on Embedded Hardware, arXiv, 2026. [Paper]
• DAVIO: Dense Monocular–Inertial SLAM with Feed-Forward Initialization and Pose-Conditioned Mapping, arXiv, 2026. [Paper] [Website] [Code]
• PROFusion: Robust and Accurate Dense Reconstruction via Camera Pose Regression and Optimization, ICRA, 2026. [Paper] [Code]
• World SLAM Model: Joint World Modeling for SLAM and Navigation, arXiv, 2026. [Paper] [Code] [Website]
• StreamRig: Exploiting Intra-Rig Geometry for Streaming Multi-Camera Odometry, arXiv, 2026. [Paper] [Code] [Website]
• CLoSeR: Closing the Loop for Long-Context Streaming Reconstruction, NeurIPS, 2026. [Paper] [Code]
• Pow3R-SLAM: Real-Time RGB-D SLAM with 3D Reconstruction Priors, arXiv, 2026. [Paper] [Code] [Website]

Calibration

• Calib3R: A 3D Foundation Model for Multi-Camera to Robot Calibration and 3D Metric-Scaled Scene Reconstruction, arXiv 2025. [Paper]

Loop Closure Detection & Place Recognition

• Multi-modal Loop Closure Detection with Foundation Models in Severely Unstructured Environments, arXiv 2025. [Paper]
• VGGT-MPR: VGGT-Enhanced Multimodal Place Recognition in Autonomous Driving Environments, arXiv 2026. [Paper]
• UniPR-3D: Towards Universal Visual Place Recognition with Visual Geometry Grounded Transformer, arXiv 2025. [Paper] [Code]
• Reloc-VGGT: Visual Re-localization with Geometry Grounded Transformer, arXiv 2025. [Paper] [Code]

Downstream Tasks

• SpatialLM: Training Large Language Models for Structured Indoor Modeling, NIPS 2025. [Paper] [Website] [Code]
• SpatialAnt: Autonomous Zero-Shot Robot Navigation via Active Scene Reconstruction and Visual Anticipation, arXiv 2025. [Paper] [Website] [Code]
• Drone-Based Indoor Mapping for Augmented Indoor 3D Modeling Using Neural Simultaneous Localization And Mapping (SLAM), * IEEE International Conference on Consumer Electronics (ICCE) 2026*. [Paper]