Boundary Cues for 3D Object Shape Recovery

Early work in computer vision considered a host of geometric cues for both shape reconstruction and recognition. However, since then, the vision community has focused heavily on shading cues for reconstruction, and moved towards data-driven approaches for recognition. In this paper, we reconsider these perhaps overlooked "boundary" cues (such as self occlusions and folds in a surface), as well as many other established constraints for shape reconstruction. In a variety of user studies and quantitative tasks, we evaluate how well these cues inform shape reconstruction (relative to each other) in terms of both shape quality and shape recognition. Our findings suggest many new directions for future research in shape reconstruction, such as automatic boundary cue detection and relaxing assumptions in shape from shading (e.g. orthographic projection, Lambertian surfaces).

• Publication year

  2013

• Venue

  2013 IEEE Conference on Computer Vision and Pattern Recognition

• Publication date

  2013-06-23

• Fields of study

  Mathematics, Computer Science

• Identifiers
  DOI 10.1109/CVPR.2013.281arXiv 1912.11566
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Diagnosing Error in Object Detectors

  2012cited by this paper
• Color Constancy, Intrinsic Images, and Shape Estimation

  2012influential reference
• Efficient Additive Kernels via Explicit Feature Maps

  2012cited by this paper
• Shape, albedo, and illumination from a single image of an unknown object

  2012cited by this paper
• Indoor scene segmentation using a structured light sensor

  2011cited by this paper
• Depth kernel descriptors for object recognition

  2011cited by this paper
• Efficient additive kernels via explicit feature maps

  2010cited by this paper
• Vlfeat: an open and portable library of computer vision algorithms

  2010influential reference
• Object Detection with Discriminatively Trained Part Based Models

  2010cited by this paper
• Kernel Descriptors for Visual Recognition

  2010cited by this paper
• Multiple kernels for object detection

  2009cited by this paper
• Ground truth dataset and baseline evaluations for intrinsic image algorithms

  2009cited by this paper
• Groups of Adjacent Contour Segments for Object Detection

  2008cited by this paper
• Image Classification using Random Forests and Ferns

  2007cited by this paper
• Object Recognition in the Geometric Era: A Retrospective

  2006cited by this paper
• The Image Plane

  2003cited by this paper
• Recovering Three Dimensional Shape from a Single Image of Curved Objects

  1987cited by this paper
• Interpreting line drawings of curved objects

  1986cited by this paper
• What Does the Occluding Contour Tell Us about Solid Shape?

  1984cited by this paper
• Machine Perception of Three-Dimensional Solids

  1963cited by this paper
• Mathematical Programming manuscript No. (will be inserted by the editor) Pegasos: Primal Estimated sub-GrAdient SOlver for SVM

  year unknowncited by this paper

• Interactive Occlusion Boundary Estimation through Exploitation of Synthetic Data

  2024cites this paper
• What does the Occluding Contour Tell us about Quantitative Shape?

  2023cites this paper
• Learning to Detect Semantic Boundaries with Image-Level Class Labels

  2022cites this paper
• LiteEdge: Lightweight Semantic Edge Detection Network

  2021cites this paper
• Pano3D: A Holistic Benchmark and a Solid Baseline for 360° Depth Estimation

  2021cites this paper
• Predicting Sharp and Accurate Occlusion Boundaries in Monocular Depth Estimation Using Displacement Fields

  2020cites this paper
• OSED: Object-specific edge detection

  2020cites this paper
• SDF-SRN: Learning Signed Distance 3D Object Reconstruction from Static Images

  2020cites this paper
• OASIS: A Large-Scale Dataset for Single Image 3D in the Wild

  2020cites this paper
• Silhouette Guided Point Cloud Reconstruction beyond Occlusion

  2019cites this paper
• Simultaneous Edge Alignment and Learning

  2018cites this paper
• Learning Single-view 3D Reconstruction of Objects and Scenes

  2018cites this paper
• Estimating Coarse 3D Shape and Pose from the Bounding Contour

  2017influential citation
• CASENet: Deep Category-Aware Semantic Edge Detection

  2017cites this paper
• Learning Category-Specific Deformable 3D Models for Object Reconstruction

  2017cites this paper
• Specular motion and 3D shape estimation.

  2017cites this paper
• A novel cloud surface shape estimation method based on SFS

  2016cites this paper
• Inverse Rendering Techniques for Physically Grounded Image Editing

  2015cites this paper
• RGB-D geometric features extraction and edge-based scene-SIRFS

  2015cites this paper
• Mutual illumination photometric stereo without calibration

  2015cites this paper
• Learning 3D Articulation and Deformation using 2D Images

  2015cites this paper
• Toward visual understanding of everyday object

  2015cites this paper
• High-quality hair modeling from a single portrait photo

  2015cites this paper
• Depth camera tracking with contour cues

  2015cites this paper
• Lifting Object Detection Datasets into 3D

  2015cites this paper
• When 3 D Reconstruction Meets Ubiquitous RGB-D Images

  2014cites this paper
• Unfolding an Indoor Origami World

  2014cites this paper
• Designing deep networks for surface normal estimation

  2014cites this paper
• When 3D Reconstruction Meets Ubiquitous RGB-D Images

  2014cites this paper
• Virtual view networks for object reconstruction

  2014cites this paper
• Reconstructing PASCAL VOC

  2014cites this paper