postPerspective

You are here

Technical Papers

Shape Signature

Thursday, 28 July, 10:45 am - 12:15 pm, Anaheim Convention Center, Ballroom D
Session Chair: Scott Schaefer, Texas A&M University

Construction of Manifolds via Compatible Sparse Representations

A sparse-optimization-based approach for constructing manifolds from discrete meshes, in which the local geometry of each chart is sparsely represented by a set of atom functions.

Ruimin Wang
University of Science and Technology of China

Ligang Liu
University of Science and Technology of China

Zhouwang Yang
University of Science and Technology of China

Kang Wang
University of Science and Technology of China

Wen Shan
University of Science and Technology of China

Jiansong Deng
University of Science and Technology of China

Falai Chen
University of Science and Technology of China

Intrinsic Girth Function for Shape Processing

This paper introduces the intrinsic girth function (IGF) for a 3D shape and applies it to shape description and geometric feature detection. The IGF is invariant under isometry, insensitive to mesh tessellations, and robust to surface noise.

Shi-Qing Xin
Ningbo University

Wenping Wang
University of Hong Kong

Shuangmin Chen
Ningbo University

Jieyu Zhao
Ningbo University

Zhenyu Shu
Zhejiang University, Ningbo Institute of Technology

SemanticPaint: Interactive Segmentation and Learning of 3D Worlds

SemanticPaint allows novices and experts alike to simultaneously acquire real-time 3D models of the world, segment these models interactively by simply reaching out and touching desired objects and surfaces, and from these segmentations learn to label new unseen parts of the world, all in an online manner.

Julien Valentin
Oxford University

Vibhav Vineet
Intel Labs

Ming-Ming Cheng
Nankai University

David Kim
Microsoft Research

Jamie Shotton
Microsoft Research

Pushmeet Kohli
Microsoft Research

Matthias Nießner
Stanford University

Antonio Criminisi
Microsoft Research

Shahram Izadi
Microsoft Research

Philip Torr
Oxford University

Detail-Preserving Mesh Unfolding for Non-Rigid Shape Retrieval

A deformation algorithm that separates each mesh vertex as far as possible while preserving geometric details of the original shape, achieving minimum bending with maximum geometric detail.

Yusuf Sahillioglu
Middle East Technical University

Ladislav Kavan
University of Utah