Technical Papers

Shape Collection

Monday, 11 August 9:00 AM - 10:30 AM | Vancouver Convention Centre, East Building, Ballroom B-C Session Chair: Michael Wimmer, Technische Universität Wien

Meta-Representation of Shape Families

Introducing meta-representation of a family of shapes for learning the essence of the family. The meta-representation encapsulates geometric distributions that encode relative arrangements of parts and can be used in several applications such as exploration of shape repositories, guided editing, and coupled editing.

Noa Fish
Tel Aviv University

Melinos Averkiou
University College London

Oliver van Kaick
Tel Aviv University

Olga Sorkine-Hornung
ETH Zürich

Daniel Cohen-Or
Tel Aviv University

Niloy J. Mitra
University College London

Organizing Heterogenous Scene Collection Through Contextual Focal Points

This paper introduces focal points for characterizing, comparing, and organizing collections of complex and heterogeneous data, and applies the concepts and algorithms developed to collections of 3D indoor scenes.

Kai Xu
Shenzhen Institutes of Advanced Technology, National University of Defense Technology

Rui Ma
Simon Fraser Universitty

Hao Zhang
Simon Fraser University

Chenyang Zhu
National University Of Defense Technology

Ariel Shamir
The Interdisciplinary Center Herzliya

Daniel Cohen-Or
Tel-Aviv University

Hui Huang
Shenzhen Institutes of Advanced Technology

Functional Map Networks for Analyzing and Browsing Large Shape Collections

This paper shows how to construct consistent functional map networks among heterogeneous shape collections. The resulting functional maps enable various applications including co-segmentation and shape exploration.

Qi-xing Huang
Stanford University

Fan Wang
Stanford University

Guibas Leonidas
Stanford University

Geometry and Context for Semantic Correspondences and Functionality Recognition in Man-Made 3D Shapes

In the presence of large geometric and topological variations, the context of a part within a 3D shape provides important cues for learning shape semantics. This paper proposes to model the context as structural relationships between parts and use them as cues for finding correspondences and for learning part functionality.

Hamid Laga
University of South Australia

Michela Mortara
Istituto di Matematica Applicata e Tecnologie Informatiche

Michela Spagnuolo
Istituto di Matematica Applicata e Tecnologie Informatiche

Estimating Image Depth Using Shape Collections

Introducing a 3D data-driven approach for estimating the depth of an image object.

Hao Su
Stanford Univeristy

Qixing Huang
Stanford University

Niloy J. Mitra
University College London

Yangyan Li
Stanford University

Leonidas Guibas
Stanford University