Technical Papers

Animating Characters

Tuesday, 12 August 3:45 PM - 5:15 PM | Vancouver Convention Centre, East Building, Ballroom A Session Chair: Ladislav Kavan, University of Pennsylvania

Tangible and Modular Input Device for Character Articulation

This paper presents a modular, novel mechanical device for animation authoring. The pose of the device is sensed at interactive rates, enabling the user to quickly pose characters rigged with a skeleton of arbitrary topology. The mapping between the physical device and virtual skeleton is semi-automatically computed starting from sparse user correspondences.

Alec Jacobson
ETH Zürich

Daniele Panozzo
ETH Zürich

Oliver Glauser
ETH Zürich

Cedric Pradalier
GeorgiaTech Lorraine

Otmar Hilliges
ETH Zürich

Olga Sorkine-Hornung
ETH Zürich

Real-Time Continuous Pose Recovery of Human Hands Using Convolutional Networks

This paper presents a method for real-time pose recovery of human hands from depth images. The pipeline includes offline labeled dataset generation as well as a convolutional network for pose inference.

Jonathan Tompson
New York University

Murphy Stein
New York University

Yann LeCun
New York University

Ken Perlin
New York University

Interactive Manipulation of Large-Scale Crowd Animation

A novel cage-based editing method for interactive manipulation of large-scale crowd animation. The cage encloses animated characters and supports convenient space/time manipulation. Examples demonstrate how the cage-based user interface mitigates the time and effort required to manipulate large crowd animation.

Jongmin Kim
Seoul National University

Yeongho Seol
Weta Digital

Taesoo Kwon
Hanyang University

Jehee Lee
Seoul National University

Robust and Accurate Skeletal Rigging From Mesh Sequences

This paper introduces an example-based rigging approach to automatically generate a linear-blend skinning model with skeleton structure. Using input from an animated mesh sequence, this approach can output its skeleton structure, joint positions, linear-blend skinning weights, and corresponding bone transformations.

Binh Le
University of Houston

Zhigang Deng
University of Houston

Interactive Generalized Penetration Depth Computation for Rigid and Articulated Models Using Object Norm

The first real-time algorithms to compute generalized penetration depth for rigid and articulated models, and their applications to physically based animation, grasp planning, and motion planning.

Tang Min
Ewha Womans University

Young Kim
Ewha Womans University