# InfiniTAM v3: A Framework for Large-Scale 3D Reconstruction with Loop Closure

arXiv: https://arxiv.org/abs/1708.00783 · HTML: https://mlsanity.com/paper/1708.00783/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Aug 2017.

## Built on by (most cited first, 25 of 25)

- [ViPE: Video Pose Engine for 3D Geometric Perception](https://mlsanity.com/paper/2508.10934/) (2025)
- [SimpleRecon: 3D Reconstruction Without 3D Convolutions](https://mlsanity.com/paper/2208.14743/) (2022)
- [Unsupervised Scale-consistent Depth Learning from Video](https://mlsanity.com/paper/2105.11610/) (2021)
- [RGBD GS-ICP SLAM](https://mlsanity.com/paper/2403.12550/) (2024)
- [BNV-Fusion: Dense 3D Reconstruction using Bi-level Neural Volume Fusion](https://mlsanity.com/paper/2204.01139/) (2022)
- [ManhattanSLAM: Robust Planar Tracking and Mapping Leveraging Mixture of Manhattan Frames](https://mlsanity.com/paper/2103.15068/) (2021)
- [NeuralBlox: Real-Time Neural Representation Fusion for Robust Volumetric Mapping](https://mlsanity.com/paper/2110.09415/) (2021)
- [ASH: A Modern Framework for Parallel Spatial Hashing in 3D Perception](https://mlsanity.com/paper/2110.00511/) (2021)
- [AGS-Mesh: Adaptive Gaussian Splatting and Meshing with Geometric Priors for Indoor Room Reconstruction Using Smartphones](https://mlsanity.com/paper/2411.19271/) (2024)
- [Object-Augmented RGB-D SLAM for Wide-Disparity Relocalisation](https://mlsanity.com/paper/2108.02522/) (2021)
- [FeatureSLAM: Feature-enriched 3D gaussian splatting SLAM in real time](https://mlsanity.com/paper/2601.05738/) (2026)
- [NeuV-SLAM: Fast Neural Multiresolution Voxel Optimization for RGBD Dense SLAM](https://mlsanity.com/paper/2402.02020/) (2024)
- [MSC-VO: Exploiting Manhattan and Structural Constraints for Visual Odometry](https://mlsanity.com/paper/2111.03408/) (2021)
- [Towards Energy-Efficiency by Navigating the Trilemma of Energy, Latency, and Accuracy](https://mlsanity.com/paper/2409.04018/) (2024)
- [GS4: Generalizable Sparse Splatting Semantic SLAM](https://mlsanity.com/paper/2506.06517/) (2025)
- [PointSLAM++: Robust Dense Neural Gaussian Point Cloud-based SLAM](https://mlsanity.com/paper/2601.11617/) (2026)
- [Rendering and Tracking the Directional TSDF: Modeling Surface Orientation for Coherent Maps](https://mlsanity.com/paper/2108.08115/) (2021)
- [Semantic Dense Reconstruction with Consistent Scene Segments](https://mlsanity.com/paper/2109.14821/) (2021)
- [Real-Time Hybrid Mapping of Populated Indoor Scenes using a Low-Cost Monocular UAV](https://mlsanity.com/paper/2203.02453/) (2022)
- [Efficient 3D Reconstruction, Streaming and Visualization of Static and Dynamic Scene Parts for Multi-client Live-telepresence in Large-scale Environments](https://mlsanity.com/paper/2211.14310/) (2022)
- [Rendering the Directional TSDF for Tracking and Multi-Sensor Registration with Point-To-Plane Scale ICP](https://mlsanity.com/paper/2301.12796/) (2023)
- [HVOFusion: Incremental Mesh Reconstruction Using Hybrid Voxel Octree](https://mlsanity.com/paper/2404.17974/) (2024)
- [Hologram: Realtime Holographic Overlays via LiDAR Augmented Reconstruction](https://mlsanity.com/paper/2405.07178/) (2024)
- [MobiFuse: A High-Precision On-device Depth Perception System with Multi-Data Fusion](https://mlsanity.com/paper/2412.13848/) (2024)
- [BayesFusion-SDF: Probabilistic Signed Distance Fusion with View Planning on CPU](https://mlsanity.com/paper/2602.19697/) (2026)

## Abstract (arXiv)

Volumetric models have become a popular representation for 3D scenes in recent years. One breakthrough leading to their popularity was KinectFusion, which focuses on 3D reconstruction using RGB-D sensors. However, monocular SLAM has since also been tackled with very similar approaches. Representing the reconstruction volumetrically as a TSDF leads to most of the simplicity and efficiency that can be achieved with GPU implementations of these systems. However, this representation is memory-intensive and limits applicability to small-scale reconstructions. Several avenues have been explored to overcome this. With the aim of summarizing them and providing for a fast, flexible 3D reconstruction pipeline, we propose a new, unifying framework called InfiniTAM. The idea is that steps like camera tracking, scene representation and integration of new data can easily be replaced and adapted to the user's needs. This report describes the technical implementation details of InfiniTAM v3, the third version of our InfiniTAM system. We have added various new features, as well as making numerous enhancements to the low-level code that significantly improve our camera tracking performance. The new features that we expect to be of most interest are (i) a robust camera tracking module; (ii) an implementation of Glocker et al.'s keyframe-based random ferns camera relocaliser; (iii) a novel approach to globally-consistent TSDF-based reconstruction, based on dividing the scene into rigid submaps and optimising the relative poses between them; and (iv) an implementation of Keller et al.'s surfel-based reconstruction approach.