Gamasutra

Faces & Portraits

Thursday, 28 July, 9:00 am - 10:30 am, Anaheim Convention Center, Ballroom D
Session Chair: Sylvain Paris, Adobe Systems, Inc.; Massachusetts Institute of Technology

Real-Time Facial Animation With Image-Based Dynamic Avatars

A novel image-based representation for dynamic 3D avatars that allows effective handling of various hairstyles and headwear, and can generate expressive facial animations with fine-scale details in real time.

Chen Cao
Zhejiang University

Hongzhi Wu
Zhejiang University

Yanlin Weng
Zhejiang University

Tianjia Shao
Zhejiang University

Kun Zhou
Zhejiang University

JALI: An Animator-Centric Viseme Model for Expressive Lip Synchronization

This general model for automatic lip-sync based only on a vocal performance and its transcript requires only seconds to lip-sync minutes of audio, and the quality of the result rivals data-driven and performance capture methods.

Pif Edwards
University of Toronto

Chris Landreth
University of Toronto

Eugene Fiume
University of Toronto

Karan Singh
University of Toronto

Perspective-Aware Manipulation of Portrait Photos

Introducing a method for modifying the apparent head pose or distance between camera and subject in a portrait photo. The approach fits a full-perspective camera and a parametric 3D head model, and then builds a 2D warp to approximate the effect of a desired change in 3D.

Ohad Fried
Princeton University

Eli Shechtman
Adobe Research

Dan Goldman
Adobe Research

Adam Finkelstein
Princeton University

Painting Style Transfer for Head Portraits Using Convolutional Neural Networks

This technique for transferring the painterly style of a head portrait onto another is based on recent advances in deep learning. Results show the method can handle a wide variety of styles while maintaining the integrity of the facial structures.

Ahmed Selim
Trinity College Dublin

Mohamed Elgharib
Qatar Computing Research Institute

Linda Doyle
Trinity College Dublin