← ArchivePaper2017
Emotion Challenge: Building a New Photoreal Facial Performance Pipeline for Games
Alex Smith, Sven Pohle, Wan-Chun Ma, Chongyang Ma, Xian-Chun Wu, Yanbing Chen, Etienne Danvoye, Jorge Jimenez, Sanjit Patel, Mike Sanders, Cyrus A. Wilson
End-to-end facial performance pipeline for real-time games combining performance capture, blendshape solving, and photoreal rendering.
Abstract
This work presents Emotion Challenge, a new photoreal facial performance pipeline built for games. The goal is a unified, robust, and scalable pipeline that spans actor likeness acquisition, character art, performance capture, and character animation. The authors discuss the specific challenges of building such a pipeline for games, including working within model and rig limitations, efficiently handling a large number of characters and a large volume of performances, achieving consistency with pre-rendered cinematics, and supporting a wide range of animation pipelines and game engines.
How to read this
- Category
- Production talk / pipeline: a photoreal facial performance pipeline for games
- Contributions
- A unified, robust, scalable pipeline spanning actor likeness acquisition, character art, performance capture, and character animation
- Discussion of game-specific constraints: working within rig/model limits, handling many characters and large performance volumes
- Approaches for consistency with pre-rendered cinematics and support across animation pipelines and game engines
- Context
- Builds on photoreal digital-actor work in the lineage of Alexander et al.'s Digital Emily Project, adapting it to the real-time and scale constraints of games.Builds on: The Digital Emily Project: Achieving a Photorealistic Digital Actor
- Correctness
- Studio practice, not peer-reviewed; results are production-proven but the talk emphasizes pipeline integration and trade-offs over quantitative evaluation, so claims are about workability at scale rather than measured accuracy.
- Clarity
- Accessible as an experience report; a first pass conveys the pipeline stages and challenges, with little need for a deep formulation pass.
- How to read it
- Read for the game-specific constraints and how the stages connect end to end; skim for the trade-offs relevant to your engine rather than expecting reproducible algorithms.
Built upon by
Nothing yet.
Related work
- Interactive Editing of Performance-based Facial Animation 2019 / SIGGRAPH Asia
- Premo: Powerful Character Rigging, Fast Animation 2023 / SIGGRAPH
- Creating an Actor-Specific Facial Rig from Performance Capture 2016 / DigiPro
- High Fidelity Facial Animation Capture and Retargeting with Contours 2013 / SCA
Keywords
This page summarises the entry and links to its original source. The archive never hosts or redistributes the publication itself.Show it in the full archive list →