Floorplan-Jigsaw: Jointly Estimating Scene Layout and Aligning Partial Scans

@article{Lin2019FloorplanJigsawJE,
  title={Floorplan-Jigsaw: Jointly Estimating Scene Layout and Aligning Partial Scans},
  author={C. Lin and Changjian Li and Wenping Wang},
  journal={2019 IEEE/CVF International Conference on Computer Vision (ICCV)},
  year={2019},
  pages={5673-5682}
}
  • C. Lin, C. Li, Wenping Wang
  • Published 2019
  • Computer Science
  • 2019 IEEE/CVF International Conference on Computer Vision (ICCV)
We present a novel approach to align partial 3D reconstructions which may not have substantial overlap. Using floorplan priors, our method jointly predicts a room layout and estimates the transformations from a set of partial 3D data. Unlike the existing methods relying on feature descriptors to establish correspondences, we exploit the 3D "box" structure of a typical room layout that meets the Manhattan World property. We first estimate a local layout for each partial scan separately and then… Expand
Extreme Relative Pose Network Under Hybrid Representations
TLDR
A novel RGB-D based relative pose estimation approach that is suitable for small-overlapping or non- overlapping scans and can output multiple relative poses and considerably boosts the performance of multi-scan reconstruction in few-view reconstruction settings. Expand
Plan2Scene: Converting Floorplans to 3D Scenes
TLDR
Qualitative and quantitative evaluations show that the Plan2Scene system produces realistic 3D interior models, outperforming baseline approaches on a suite of texture quality metrics and as measured by a holistic user study. Expand
Zillow Indoor Dataset: Annotated Floor Plans With 360deg Panoramas and 3D Room Layouts
TLDR
A large indoor dataset with 71,474 panoramas from 1,524 real unfurnished homes, ZInD provides annotations of 3D room layouts, 2D and 3D floor plans, panorama location in the floor plan, and locations of windows and doors, which is the largest real dataset with layout annotations. Expand
Blocks-World Cameras
TLDR
The proposed Blocks-World Cameras are a class of imaging systems which directly recover dominant planes of piece-wise planar scenes (Blocks-World), without requiring point clouds, and develop a novel geometric algorithm for recovering scene planes without explicit correspondence matching, thereby avoiding computationally intensive search or optimization routines. Expand
Robot Localization in Floor Plans Using a Room Layout Edge Extraction Network
TLDR
A memory and computationally efficient monocular camera-based localization system that allows a robot to estimate its pose given an architectural floor plan and estimates the robot pose using a particle filter that matches the extracted edges to the given floor plan is proposed. Expand

References

SHOWING 1-10 OF 54 REFERENCES
Joint Layout Estimation and Global Multi-view Registration for Indoor Reconstruction
TLDR
A novel method to jointly solve scene layout estimation and global registration problems for accurate indoor 3D reconstruction and experimentally verify the proposed method with the publicly available synthetic and real-world datasets. Expand
LayoutNet: Reconstructing the 3D Room Layout from a Single RGB Image
TLDR
An algorithm to predict room layout from a single image that generalizes across panoramas and perspective images, cuboid layouts and more general layouts (e.g. "L"-shape room) is proposed, which achieves among the best accuracy for perspective images and can handle both cuboid-shaped and moregeneral Manhattan layouts. Expand
Efficient 3D Room Shape Recovery from a Single Panorama
  • Hao Yang, H. Zhang
  • Mathematics, Computer Science
  • 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2016
TLDR
An algorithm that can automatically infer a 3D shape from a collection of partially oriented superpixel facets and line segments and is efficient, that is, the inference time for each panorama is less than 1 minute. Expand
Piecewise Planar and Compact Floorplan Reconstruction from Images
TLDR
This paper presents a system to reconstruct piecewise planar and compact floorplans from images, which are then converted to high quality texture-mapped models for free- viewpoint visualization, and shows that the texture mapped mesh models provide compelling free-viewpoint visualization experiences, when compared against the state-of-the-art and ground truth. Expand
Exploiting 2D Floorplan for Building-Scale Panorama RGBD Alignment
TLDR
This paper presents a novel algorithm that utilizes a 2D floorplan to align panorama RGBD scans and formulate a novel Markov Random Field inference problem as a scan placement over the floorplan, as opposed to the conventional scan-to-scan alignment. Expand
Structured Indoor Modeling
TLDR
A novel 3D modeling framework that reconstructs an indoor scene as a structured model from panorama RGBD images and proposes a new room segmentation algorithm and an offset-map reconstruction algorithm that can enforce architectural shape priors far beyond existing state-of-the-art. Expand
Box in the Box: Joint 3D Layout and Object Reasoning from Single Images
TLDR
A branch and bound algorithm which is guaranteed to retrieve the global optimum of the joint problem and a new decomposition method, which generalizes integral geometry to triangular shapes, and allows us to bound the different terms in constant time. Expand
Recovering the spatial layout of cluttered rooms
TLDR
This paper introduces a structured learning algorithm that chooses the set of parameters to minimize error, based on global perspective cues, and gains robustness to clutter by modeling the global room space with a parameteric 3D “box” and by iteratively localizing clutter and refitting the box. Expand
DPPTAM: Dense piecewise planar tracking and mapping from a monocular sequence
  • Alejo Concha, Javier Civera
  • Computer Science
  • 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
  • 2015
TLDR
This paper proposes a direct monocular SLAM algorithm that estimates a dense reconstruction of a scene in real-time on a CPU, based on the information of a superpixel segmentation and the semidense map from highly textured areas. Expand
Efficient structured prediction for 3D indoor scene understanding
TLDR
It is shown that the potentials commonly used in the literature can be decomposed into pair-wise potentials by extending the concept of integral images to geometry and results in large improvements in performance over the state-of-the-art, while being orders of magnitude faster. Expand
...
1
2
3
4
5
...