KinectFusion: Real-time dense surface mapping and tracking

By Richard A. Newcombe, Shahram Izadi, Otmar Hilliges, David Molyneaux, David Kim, Andrew J. Davison, Pushmeet Kohli, Jamie Shotton, Steve Hodges, Andrew William Fitzgibbon. 2011

4,655 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Geology, 3D Vision & Pose, Robotics & Control

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

In Geology

In 3D Vision & Pose

In Robotics & Control

Citations in the corpus

ScanNet: Richly-Annotated 3D Reconstructions of Indoor Scenes (2017, neutral): “The RGB-D scanning system must be trivial to use, the data processing must be robust and automatic, the semantic annotations must be crowdsourced, and the ﬂow of data through the system must be handled by a tracking server.”
OctNet: Learning Deep 3D Representations at High Resolutions (2016, background): “ese cases, the input to the network is of two-dimensional nature, e.g., in image classiﬁcation [18], object detection [35] or semantic segmentation [13]. However, recent advances in 3D reconstruction [33] and graphics [21] allow capturing and modeling large amounts of 3D data. At the same time, larg”
DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation (2019, methodology): “This inherits from the success of fusion approaches that utilize a truncated SDF (TSDF), pioneered in [15, 37], to combine noisy depth maps into a single 3D model.”
OctNetFusion: Learning Depth Fusion from Data (2017, methodology): “[31] showed that real-time 3D modeling is feasible using this approach.”
Shape Completion Using 3D-Encoder-Predictor CNNs and Shape Synthesis (2016, background): “Since the introduction of commodity range sensors such as the Microsoft Kinect, RGB-D scanning has gained a huge momentum in both ofﬂine and real-time contexts [26, 3, 28, 42, 4, 7].”
RLBench: The Robot Learning Benchmark & Learning Environment (2019, background): “However, if we would like a manipulation system to make use of a SLAM map, it is not currently clear what the best way to represent this map is: whether it be sparse [54], [55], dense [56], [57], or semi-dense [58].”
SUN3D: A Database of Big Spaces Reconstructed Using SfM and Object Labels (2013, methodology): “Instead (Figure 8), we improve the depth map using a Truncated Signed Distance Function (TSDF) [15] to voxelize the space, accumulating the depth map from nearby frames (e.”
A Large Dataset of Object Scans (2016, background): “where np is the surface normal at p and K is a set of corresponding point pairs found by projective data association [5].”
VConv-DAE: Deep Volumetric Shape Learning Without Object Labels (2016, background): “To overcome some of these difficulties, there is a large body of work on fusing multiple scans into a single 3D model [2].”
Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age (2016, positive): “attracting increasing attention after the seminal work [183].”
A benchmark for RGB-D visual odometry, 3D reconstruction and SLAM (2014, methodology): “We evaluate 5 different odometers; (i) DVO [13] which performs full frame alignment between pairs of RGB-D frames making use of a robust cost function to improve robustness to outliers; (ii) FOVIS [14] which uses FAST feature correspondences between pairs of RGB-D frames to resolve camera motion, re”
The Replica Dataset: A Digital Replica of Indoor Spaces (2019, methodology): “Given the 6 DoF poses from the SLAM system, depth images are fused into a truncated signed distance function (TSDF) akin to KinectFusion [23].”
Deformable Shape Completion with Graph Convolutional Autoencoders (2017, background): “For static scenes, this problem was explored extensively, e.g., in [37, 38, 15].”
Learning 3D Object Categories by Looking Around Them (2017, methodology): “We extract this information using off-theshelf methods: the structure-from-motion (SFM) algorithm COLMAP for RGB sequences [33, 34], and an open-source implementation [31] of KinectFusion (KF) [24] for RGBD sequences.”
Beyond Point Clouds: Scene Understanding by Reasoning Geometry and Physics (2013, methodology): “With the recent development of Kinect camera and the SLAM techniques, there has been growing interest in studying these properties in the literature [17].”
MaskFusion: Real-Time Recognition, Tracking and Reconstruction of Multiple Moving Objects (2018, background): “KinectFusion [33] proved that a truncated signed distance function (TSDF) based map representation can achieve fast and robust mapping and tracking in small environments.”
DeepContext: Context-Encoding Neural Pathways for 3D Holistic Scene Understanding (2016, methodology): “2We use a 128× 128× 64 grid for the TSDF of a whole scene, with a voxel unit size of 0.05 meters and the truncation set at 0.15 meters. office area, on the other hand, usually includes a computer on a desk with a nearby chair.”