Database‐Assisted Object Retrieval for Real‐Time 3D Reconstruction

By Yangyan Li, Angela Dai, Leonidas J. Guibas, Matthias Nießner. 2015

228 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computational Mechanics, Geology, 3D Vision & Pose

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 Computational Mechanics

In Geology

In 3D Vision & Pose

Citations in the corpus

Learning Shape Priors for Single-View 3D Completion and Reconstruction (2018, background): “…our ShapeHD aims to model ambiguity on the next level: even among the realistic shapes, there are still multiple shapes explaining the observation well (c). growing number of papers on exploiting shape structures and regularities [34,51], and papers on leveraging strong database priors [46,32,4].”
Shape Completion Using 3D-Encoder-Predictor CNNs and Shape Synthesis (2016, methodology): “Another idea is to ﬁnd identical CAD models in a shape database for a given partial input shape and align it with the scan [24, 33, 16, 21, 35].”
Mask2CAD: 3D Shape Prediction by Learning to Segment and Retrieve (2020, methodology): “Most methods focus on aligning CAD models to RGB-D scan, point cloud geometry, or 2D-3D surface mapping though various geometric feature matching techniques [1,13,19,21,24,37].”
Deep Hough Voting for 3D Object Detection in Point Clouds (2019, methodology): “Examples include: [27] where a pair-wise semantic context potential helps guide the proposals objectness score; templatebased methods [26, 32, 28]; Sliding-shapes [41] and its deep learning-based successor [42]; Clouds of Oriented Gradients (COG) [38]; and the recent 3D-SIS [12].”
Data-driven structural priors for shape completion (2015, methodology): “Other approaches take advantage of efficient featurebased descriptors and simply use the most similar object to complete the 3D scan [Nan et al. 2012; Kim et al. 2013b; Li et al. 2015].”
3D Object Reconstruction from a Single Depth View with Adversarial Learning (2017, methodology): “Given a partial shape input, [25] [29] [36] [45] [47] try to retrieve an identical or most likely CAD model and align it with the partial scan.”
Back to Reality: Weakly-supervised 3D Object Detection with Shape-guided Label Enhancement (2022, methodology): “3D Object Detection: Early 3D object detection methods mainly include template-based methods [16, 18, 23] and sliding-window methods [37, 38].”
3D Object Dense Reconstruction from a Single Depth View (2018, unknown): “Given a partial shape input, [31] [32] [33] [34] [35] [36] try to retrieve an identical or most likely 3D model and align it with the partial scan.”
ImVoteNet: Boosting 3D Object Detection in Point Clouds With Image Votes (2020, methodology): “To locate objects using purely geometric information, one popular line of methods is based on template matching using a collection of clean CAD models either directly [21, 28, 25], or through extracted features [44, 2].”
3DMatch: Learning the Matching of Local 3D Geometry in Range Scans (2016, methodology): “[20], we use an implementation provided by the authors.”
Retrieval-Enhanced Machine Learning (2022, unknown): “[20, 35, 40, 54].”
Shape completion enabled robotic grasping (2016, methodology): “s Random Forests. In addition, work by [13] uses an exemplar based approach for the same task. Others [14][15] have developed algorithms to learn 3D occupancy grids directly from 2D images. Li et. al [16] uses a database of pre-existing models to do completion. Depth Map X Convert to Occupancy Grid ”
Scan2CAD: Learning CAD Model Alignment in RGB-D Scans (2018, methodology): “[25] propose a formulation based on a hand-crafted TSDF feature descriptor to align CAD models in real-time to RGB-D scans.”
ScanComplete: Large-Scale Scene Completion and Semantic Segmentation for 3D Scans (2017, unknown): “One can also ‘complete’ shapes by replacing scanned geometry with aligned CAD models retrieved from a database [20, 32, 15, 17, 33].”
PCN: Point Completion Network (2018, background): “Some [15, 24, 32, 34, 43] retrieve the complete shape directly while some [17, 19, 28, 45, 52] retrieve object parts and then assemble them to obtain the complete shape.”