Digital Face Cloning - PowerPoint PPT Presentation

About This Presentation
Title:

Digital Face Cloning

Description:

George Borshukov J.P. Lewis Lance Williams. Paul Debevec Fred Pighin Li Zhang. David Bennett Christophe Hery Steve Sullivan ... 1998: Hollywood is ready ' ... – PowerPoint PPT presentation

Number of Views:243
Avg rating:3.0/5.0
Slides: 19
Provided by: graphics80
Category:

less

Transcript and Presenter's Notes

Title: Digital Face Cloning


1
Digital Face Cloning
  • David Bennett Christophe Hery Steve Sullivan
  • George Borshukov J.P. Lewis Lance Williams
  • Paul Debevec Fred Pighin Li Zhang

2
Boston
Seattle
San Francisco
Los Angeles
  • David Bennett Christophe Hery Steve Sullivan
  • George Borshukov J.P. Lewis Lance Williams
  • Paul Debevec Fred Pighin Li Zhang

3
Definition
  • Process of capturing an actors performance
    and/or likeness in a digital face model

4
E.g. photogrammetry (input)
5
E.g. photogrammetry (output)
6
Motivation
  • Replicate a live actor (e.g. celebrity)
  • Stunts
  • Extreme makeup
  • Younger/older/slimmer clone
  • Creating realistic faces from first principles is
    difficult (digitizing is easier than modeling)

7
Purpose of the course
  1. Review the field
  2. Suggest directions

8
Digital Clones History
9
CG Humans in the movies
  • First 3D actor representation in the movies
    Peter Fonda in FutureWorld (1976!)
  • Not intended to be photorealistic

10
1990s
11
1990s
  • Terminator 2 (brief, texture projection)
  • Jurrasic Park (legs only, does not count)
  • Titanic (distant)
  • Startup attempt Virtual Celebrity/Global Icons,
    purchased virtual rights to Sammy Davis Jr.,
    others

12
1998 Hollywood is ready
  • "I don't see them as competing with real actors,
    Richard Masur, head of Screen Actors Guild
  • Projected uses de-aging, celebrity vacuum ads.
  • A television series with a virtual cast was
    considered

13
Recent and current
14
Recent and current
  • Space Cowboys (2000), Enemy at the Gates (2001)
    brief and distant, but successful
  • Disney Facial Cloning test (afternoon session)
  • Matrix sequels (2003) superpunch full-screen
    for 13 seconds (afternoon session)
  • SpiderMan 2 (reflectance capture)
  • Lemony Snicket CG baby in several shots
    (afternoon session)
  • Discovery Channel Churchill/Hitler clones (MPC)

15
Observation
  • There has not been a breakthrough introduction
    of digital clones
  • Rather, clones were initially used in the
    distance, and briefly they are getting closer,
    and withstanding longer study.
  • Current state of the art full screen for
    several seconds (see afternoon Perception session)

16
Syllabus morning (theory)
  • 830 Introduction and Overview Fred Pighin
  • 900 Face Scanning and Dense Motion Capture
    Technologies Li Zhang
  • 1000 Break or Mathematical Background (schedule
    change)
  • 1015 Reflectance Modeling and Capture Paul
    Debevec and J.P. Lewis
  • 1115 Facial Parameterization and Cross-Mapping
    J.P. Lewis and Fred Pighin
  • 1215 Lunch

17
Syllabus afternoon (practice)
  • 130 Case study Face Cloning at ILM Steve
    Sullivan and Christophe Hery
  • 215 Perception Experiment (schedule change)
  • 230 Case Study Disney Facial Cloning Lance
    Williams
  • 315 Case study Leaping the Uncanny Valley with
    Data   (Face Cloning in the Matrix sequels)
    George Borshukov
  • 400 Break
  • 415 Case study Polar Express David Bennet
  • 500 Perception of Facial Realism J.P. Lewis
  • 515 Panel on the Future of Digital Face Cloning
    All

18
Face Cloning at Siggraph
  • Courses
  • 10. Realistic Materials in Computer Graphics
  • 28. From Mocap to Movie The Making of "The Polar
    Express
  • 29. High-Dynamic-Range Imaging and Image-Based
    Lighting
  • Papers (Skin Faces session on Monday)
  • Automatic Determination of Facial Muscle
    Activations From Sparse Motion Capture Marker
    Data
  • Face Transfer With Multilinear Models
  • Postproduction Relighting and Reflectance
    Transformation with Time-Multiplexed Illumination
  • Sketches
  • Image-Based Rendering From a Sparse Set of Images
  • Implementation of Modeling Hair From Multiple
    Views
  • Model Flowing Capturing and Tracking of
    Deformable Geometry
  • Performance Geometry Capture for Spatially
    Varying Relighting
  • Joint Motion and Reflectance Capture for
    Relightable 3D Video
Write a Comment
User Comments (0)
About PowerShow.com