Technical Papers

Depth for All Occasions

Thursday, 14 August 2:00 PM - 3:30 PM | Vancouver Convention Centre, East Building, Ballroom B-C Session Chair: Johannes Kopf, Microsoft Research

Color Map Optimization for 3D Reconstruction With Consumer Depth Cameras

A global optimization approach for color mapping three-dimensional reconstructions obtained from handheld RGB-D scans. The approach optimizes camera poses in tandem with non-rigid correction functions for all images. All parameters are optimized jointly to maximize the photometric consistency of the reconstructed color map.

Qian-Yi Zhou
Stanford University

Vladlen Koltun
Adobe Research

Real-Time Non-Rigid Reconstruction Using an RGB-D Camera

A novel and comprehensive solution for markerless reconstruction of non-rigidly deforming physical objects with arbitrary shape in real time. The algorithm is an order of magnitude faster than state-of-the-art methods, while matching the quality and robustness of many off-line algorithms.

Michael Zollhöfer
Friedrich-Alexander-Universität Erlangen-Nürnberg

Matthias Nießner
Stanford University

Shahram Izadi
Microsoft Research Cambridge

Christoph Rehmann
Microsoft Research Cambridge

Christopher Zach
Microsoft Research Cambridge

Matthew Fisher
Stanford University

Chenglei Wu
Max-Planck-Institut für Informatik

Andrew Fitzgibbon
Microsoft Research Cambridge

Charles Loop
Microsoft Research

Christian Theobalt
Max-Planck-Institut für Informatik

Marc Stamminger
Friedrich-Alexander-Universität Erlangen-Nürnberg

Proactive 3D Scanning of Inaccessible Parts

Proactive scanning is a paradigm that allows the user to modify the scene while scanning it, in order to access occluded regions. The technique utilizes a registration algorithm to compute motion trajectories and separate between motions in the scene, then reconstructs the static parts into a complete unified 3D model.

Feilong Yan
Shenzhen VisuCA Key Lab, Shenzhen Institutes of Advanced Technology

Andrei Sharf
Ben-Gurion University

Wenzhen Lin
Shenzhen VisuCA Key Lab, Shenzhen Institutes of Advanced Technology

Hui Huang
Shenzhen VisuCA Key Lab, Shenzhen Institutes of Advanced Technology

Baoquan Chen
Shandong University

Coupled Structure-From-Motion and 3D Symmetry Detection for Urban Façades

Dominant repeated structures in building façades lead to ambiguity in establishing correspondences across images. This paper presents a coupled structure-from-motion and symmetry detection framework to resolve this ambiguity.

Duygu Ceylan
École polytechnique fédérale de Lausanne

Niloy J. Mitra
University College London

Youyi Zheng
Yale University

Mark Pauly
École polytechnique fédérale de Lausanne

Painting-to-3D Model Alignment via Discriminative Visual Elements

This paper describes a technique that finds an approximate viewpoint of arbitrary 2D depictions of an architectural site with a 3D model of the site and shows reliable alignment despite large variations in
rendering style (watercolor, sketch, painting, historical photograph) and structural changes of the scene.

Mathieu Aubry
INRIA, Technische Universität München

Bryan Russell
Intel Corporation

Josef Sivic
INRIA-ENS