Shape Analysis

Technical Papers

Shape Analysis

Tuesday, 11 August 3:45 PM - 5:35 PM | Los Angeles Convention Center, Room 153A-C Session Chair: Vladimir Kim, Stanford University

Interaction Context (ICON): Towards a Geometric Functionality Descriptor

Introducing a contextual descriptor that aims to provide a geometric description of the functionality of a 3D object. The descriptor, called interaction context (ICON), explicitly captures object-to-object interactions in the context of a given scene.

Ruizhen Hu
Shenzhen Institute of Advanced Technology, Zhejiang University, Simon Fraser University

Chenyang Zhu
Simon Fraser University

Oliver van Kaick
Carleton University

Ligang Liu
University of Science and Technology of China

Ariel Shamir
Interdisciplinary Center Herzliya

Hao (Richard) Zhang
Simon Fraser University

Elements of Style: Learning Perceptual Shape Style Similarity

The human perception of stylistic similarity transcends structure and function. For instance, a bed and a dresser may share a common style. This paper introduces and successfully validates the first structure-transcending style-similarity measure for shapes, one closely aligned with the human perception of style similarity.

Zhaoliang Lun
University of Massachusetts Amherst

Evangelos Kalogerakis
University of Massachusetts Amherst

Alla Sheffer
The University of British Columbia

Style Compatibility for 3D Furniture Models

This method for learning to predict the stylistic compatibility between 3D furniture models from different object classes introduces part-aware geometric features and learns class-specific mappings from them to a shared feature space. The method is effective at predicting style compatibility, and it enables many applications.

Tianqiang Liu
Princeton University

Aaron Hertzmann
Adobe Systems Incorporated

Wilmot Li
Adobe Systems Incorporated.

Thomas Funkhouser
Princeton University

Semantic Shape Editing Using Deformation Handles

With this shape-editing method, users edit shapes with a set of semantic attributes, thus avoiding the need for detailed geometric manipulations. The method provides a platform for quick design explorations and allows non-experts to produce semantically guided shape variations that are otherwise difficult to attain.

Mehmet Ersin Yumer
Carnegie Mellon University

Siddhartha Chaudhuri
Cornell University

Jessica Hodgins
Carnegie Mellon University

Levent Burak Kara
Carnegie Mellon University

Single-View Reconstruction via Joint Analysis of Image and Shape Collections

In this approach to automatic 3D reconstruction of objects from images, the key idea is to jointly analyze a collection of images of different objects along with a smaller collection of available 3D models. The images are analyzed and reconstructed together.

Qi-xing Huang
Toyota Technological Institute at Chicago

Hai Wang
Toyota Technological Institute at Chicago

Vladlen Koltun
Intel Labs