SALVe: Semantic Alignment Verification for Floorplan Reconstruction from Sparse Panoramas
Not on the map. Why?
Abstract (arXiv)
We propose a new system for automatic 2D floorplan reconstruction that is enabled by SALVe, our novel pairwise learned alignment verifier. The inputs to our system are sparsely located 360^circ panoramas, whose semantic features (windows, doors, and openings) are inferred and used to hypothesize pairwise room adjacency or overlap. SALVe initializes a pose graph, which is subsequently optimized using GTSAM. Once the room poses are computed, room layouts are inferred using HorizonNet, and the floorplan is constructed by stitching the most confident layout boundaries. We validate our system qualitatively and quantitatively as well as through ablation studies, showing that it outperforms state-of-the-art SfM systems in completeness by over 200%, without sacrificing accuracy. Our results point to the significance of our work: poses of 81% of panoramas are localized in the first 2 connected components (CCs), and 89% in the first 3 CCs. Code and models are publicly available at https://github.com/zillow/salve.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 26 of its 69 references that are in the map. First posted on arXiv 27 Jun 2024.
Builds on
26 of 69 references.
References · 26
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2021Training Verifiers to Solve Math Word Problemscited by 10,109
- 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,443
- 2017Matterport3D: Learning from RGB-D Data in Indoor Environmentscited by 1,376
- 2021LoFTR: Detector-Free Local Feature Matching with Transformerscited by 984
- 2016Semantic Scene Completion from a Single Depth Imagecited by 472
- 2019Structured3D: A Large Photo-realistic Dataset for Structured 3D Modelingcited by 286
- 2017A Survey of Structure from Motioncited by 158
- 2019HorizonNet: Learning Room Layout with 1D Representation and Pano Stretch Data Augmentationcited by 108
- 2020HoHoNet: 360 Indoor Holistic Understanding with Latent Horizontal Featurescited by 105
- 2020House-GAN: Relational Generative Adversarial Networks for Graph-constrained House Layout Generationcited by 87
- 2021Generate & Rank: A Multi-task Framework for Math Word Problemscited by 82
14 more
- 2017Camera Relocalization by Computing Pairwise Relative Poses Using Convolutional Neural Networkcited by 79
- 2021Wide-Baseline Relative Camera Pose Estimation with Directional Learningcited by 49
- 2021Planar Surface Reconstruction from Sparse Viewscited by 40
- 2021Pano3D: A Holistic Benchmark and a Solid Baseline for 360o Depth Estimationcited by 37
- 2019Floor-SP: Inverse CAD for Floorplans by Sequential Room-wise Shortest Pathcited by 36
- 2018FloorNet: A Unified Framework for Floorplan Reconstruction from 3D Scanscited by 35
- 2020Audio-Visual Floorplan Reconstructioncited by 32
- 2021MonteFloor: Extending MCTS for Reconstructing Accurate Large-Scale Floor Planscited by 26
- 2019Conv-MPN: Convolutional Message Passing Neural Network for Structured Outdoor Architecture Reconstructioncited by 25
- 2022Trust, but Verify: Cross-Modality Fusion for HD Map Change Detectioncited by 25
- 2018Extreme Relative Pose Estimation for RGB-D Scans via Scene Completioncited by 22
- 2019Extreme Relative Pose Network under Hybrid Representationscited by 19
- 2018Floorplan-Jigsaw: Jointly Estimating Scene Layout and Aligning Partial Scanscited by 11
- 2023Distributed Global Structure-from-Motion with a Deep Front-Endcited by 8
Built on by
Citing papers per year: 2025: 1; 2026: 1.
2 papers.