Title: Interactive Media Group iMG
1Interactive Media GroupiMG
Ali Arya Carleton School of Information
Technology Department of Systems and Computer
Engineering
2CSIT
- Carleton School of Information Technology
- Areas of research
- Networking
- Multimedia
- Independent undergrad program
- Graduate program through Dept. of Systems and
Computer Engineering
3iMG _at_ SCE
- Interactive Media Group (iMG)
- Part of Department of Systems and Computer
Engineering (SCE) - Motivations new multimedia systems and
interactive technologies - Computer games
- Interactive virtual environments
- Interactive web applications
- New human-computer interfaces
- New media and digital art
4Areas of Research
- Multimedia Systems and Applications
- Human-Computer Interaction
- Computer Graphics and Animation
- Virtual Worlds and Virtual Characters
- Virtual and Digital Art
5Some Research Projects
- iFACE, facial animation engine
- SPICE, interactive virtual dance
- Interactive Story Spaces
- Digital Libraries
- Medical Planning
- Game and Simulation
- Automotive Displays and Multimodal Interaction
6img.csit.carleton.ca
7Animated Images
8Milestones
- Disney studios
- Cel animation
- Computer generated animation in early 60s
- Mr. Computer Image ABC with Scanimate
- 1970s
- Voyager 2 by JPL (James Blinn)
- Star Wars by ILM (George Lucas)
9Pong by Atari (1972)
- The story begins.
- The first commercially successful video game
- Table tennis
10Games to Remember
11Commodore 64 - 1982
12Computer Games
- computer game
- a game composed of a computer-controlled virtual
universe that players may interact with in order
to achieve a goal (or set of goals) - video game
- a computer game where a video display is the
primary feedback device
13Gaming Platforms
- PC
- Windows
- Mac
- Console
- Commodore 64 and Atari ?
- X-Box from Microsoft
- PlayStation from Sony
- Game Boy from Nintendo
- Arcade
14Milestones
- 1980s
- Disney TRON combined live action with 3D computer
animation - Atari and video games
- Visual effects, Star Wars
- Genesis Effect in Star Trek II by ILM
- Pixar RenderMan
- Little Mermaid (89) the last traditional film
15Geris Game by RenderMan
16Milestones
- Early 1990s
- Terminator II (91) by ILM
- 1st to include morphing and human simulation
- Jurassic Park (93) by ILM
- Babylon 5 (93)
- 1st 3D animation, made by Amiga and Mac
- Cocacola Polar Bear (93) commercial
- Disney CAPS, Lion King
17Coca-cola Polar Bear
18Doom by id Software (1993)
- Pioneer in 3D graphics
- Next generation of Wolfenstein 3D
- Re-usable engine
- Violence, satanic imagery, and other
controversies
19Milestones
- Late 1990s
- Toy Story (95)
- 1st computer generated feature film
- DreamWorks vs. Disney/Pixar
- Antz, Prince of Egypt
- Bugs Life, Mulan, Toy Story 2
- Fantasia 2000 for IMAX
- Armagedon (98)
20Feature Films
21Game Genres
- Role-playing games (RPG)
- Third-person shooter (TPS)
- First-person shooter (FPS)
- Sports and racing
- Simulation (e.g. flight)
- Strategy
-
22Flight Simulators
23Racing Game
24FPS Games
25The Sims by Maxis (2000)
- Simulated characters, simulated life
- SimCity (1989)
- God-game!
26Multimedia Webpage
27Milestones
21st Century
28Interactive Webpage
29Social Agents
- Virtual characters
- Interactive
- Social
- Realistic or stylistic
- Storyteller
- iFACE Demo
- Arya and DiPaola, 2005
- A little bit of history
308th Art
- Ricciotto Canudo (1879-1923)
- Seventh Art (Cinema)
- Three arts of space - painting, architecture and
dance - Three arts of time - music, theatre, and
literature - Multimedia, the 8th art!
- One art of space photography
- Two arts of time - music and literature
- One art of time and space and light - cinema
31Interactive Multimedia Presentation
- Live-action vs. animation
- Sequential vs. interactive
- Merging of different domain into one medium
- Interactivity
- Face-based communication
- Verbal and non-verbal
- Social user experience
32Characters
33Making Faces ?
- Virtual Characters Substitute for real human
figures in a social interface - Virtual Environments
- Training and Customer Service
- Games
- Special Effects
- Video Phones
34A Social User Interface
35Another Example of SUI
36Social Communication
- Creating a social exchange
- Natural aspects of human communication
- Verbal cues
- Gestures
- Facial expressions
- Communication is a major functionality of
computer-based systems - Explicitly or implicitly
37Social Conversational Agent
- Flexibility and availability of virtual
characters vs. live-action - Avatar as a visual agent
- Interactivity and understandability
- As opposed to traditional UI objects
- Front end
- Realistic or stylistic character
- Back end
- Intelligence, rules of interaction, scenario
38Face Multimedia Object(Arya and DiPaola, 2004)
- Object-based multimedia standards
- MPEG-4
- Media types
- Audio
- Video
-
- Face-centric applications
- Face media type
39FMO Advantages
- Encapsulation of face functionality
- Coordination of face-related functions
- Ease of application development
- Through dedicated API and scripts
- Independence of characters from other objects and
media type - Autonomous behavior
- 3D and 2D with the same interface
- Hierarchical Modularity
40Computer Facial Animation
- Fredric Parke, early 1970s, first computer model
for face - Some applications
- Computer games
- Social conversational agents
- Visual communication
- Video phones, etc
- Visual effects in movies
41Geris Game by RenderMan
42Final Fantasy ?
43Modeling/Animation Methods
- 2D
- Morphing and image transforms
- No 3D computation, usually more realistic, but
might need a database of images - 3D
- Direct parameterized
- Physically-based (skin, muscle, bone)
44Morphing
Mapping of each pixel is a function of the
mappings of some Control Points. Needs manual
setting of controls.
45Talking Head (Ezzat et al.)
Optical Flow The mapping of each pixel of source
image is calculated based on the similar point
found in target images. Works for one set of
images (one person, one view) only.
46Feature-based Image Transformation (ShowFace,
Arya et al., 2002)
T (Fs ,M) M Ft - Fs
Learning
Apply T Warp image
Runtime
473D Faces
- Models
- Physically-based
- Parameterized
- Performance-based
- Data sources
- 3D Digitizer
- Laser scanner
- Photogrammetry
483D Digitized Model
49Scanned Range Data
50Scanned Colour Data
51Adapted Mesh
523D From 2D
533D Face Models (Virtual Human Director, Lee et
al.)
54FACS
- Facial Action Coding System (FACS)
- 64 basic facial Action Units (AUs)
- Inner Brow Raiser
- Mouth Stretch
- Head Turn Left
-
- Originally not for computer graphics and
animation, but used extensively
55MPEG-4
- Motion Picture Experts Group
- Standards
- MPEG-1 (standard audio/video compression)
- MP3 (MPEG-1 Layer-3)
- MPEG-2 (streaming)
- MPEG-4 (object-based multimedia)
- MPEG-7 (content description)
56iFACE System (Arya and DiPaola, 2004)
- Interactive Face Animation Comprehensive
Environment - Face engine
- Modular multi-layer framework
- Authoring and runtime components
- PC version
- .NET framework
- C
- DirectX
- Console version? Mobile version?
57iFACE Architecture
Application Layer (GUI App, Web Page, etc)
Wrapper Layer (Form Controls, ActiveX, etc)
Stream Layer (Frame Reader/Writer, Mixer, etc)
Data Layer (Geometry, Behaviour, Media)
58iFACE Parameter Spaces
Events
Knowledge
Geometry
Mood
Personality
59iFACE Geometry
60Different Geometry Types
- Layered interfaces allow working with different
data types in a similar way.
61Face Regions
62Believability vs. Realism
- Disney animation makes audience really believe
in characters. There is a special ingredient in
our type of animation that produces drawing that
appear to think and make decisions and act of
their own volition it is what creates the
illusion of life. - F. Thomas and O. Johnston
63Believability
- Believability. That is what we were striving
for. - Chuck Jones (Warner Bros.)
- From the earliest days, it has been the
portrayal of emotions that has given the Disney
characters the illusion of life. - Thomas and Johnston (Disney)
64iFACE Behaviour (Arya DiPaola, 2005)
- Knowledge
- Scenario (character-independent actions)
- Rules of interaction
- Personality
- Long-term and character-dependent
- Mood
- Short-term emotional states
65Emotions
- Charles Darwin
- Universality of emotions
- Expression of the Emotions in Men and Animals,
1872 - Paul Ekman and W.V. Friesen
- Analysis of facial expressions in different
cultures - Facial Action Coding System (FACS), 1978
66FACS Sample Action Units
67Universal Emotions
- Sadness and agony
- Joy
- Anger
- Surprise
- Fear
- Disgust
- Contempt
68Facial Expressionssee Paul Ekman, Emotions
Revealed, 2003
69Homers Expressions
Clueless Annoyed Playful Pleased Stupified Mad
about Boy
70Facial Action for Expressions
- Joy
- eye-lid tightened
- cheeks raised
- brows slightly lowered
- wrinkles around eyes (especially corners)
- Sadness
- upper eye-lid lowered
- lower eye-lid raised
- inner brows raised
- lip cornered lowered
- lower lip raised slightly
- cheeks raised
71Russells Mood Model
72Expression Units
- Research on Expression Units (EUs)
- Collective set of all small facial actions
involved in expressions - Associate expression units to parameters
(Russells dimensions) - Create different expressions directly from
parameters and EUs
73Expression Units Example
Exp-1, average, Exp-2, blend using EUs
74Mood-dependency
- Moods affect the way facial actions are performed.
75Do We Really Depend on the Kindness of
Strangers?
76Static Visual Cues
- Standard Moods
- Joy
- Sadness
- Anger
- Fear
- Disgust
- Surprise
- Custom-defined Expressions
- Eye, brow, mouth and head rest position
- Speaking out of a corner of mouth
- Gaze (looking into camera)
- Gender
- Age
- General Appearance
- Round face, Full lips, Eye separation, Nose
shape, Brow thickness, etc - Baby face vs. mature
- Other attraction-related features
77Dynamic Visual Cues
- 3D Head Movements
- Frequency
- Duration
- Direction (yaw, pitch, roll)
- Nodding (especially in emphasis for speech)
- Laughing
- Raising eyebrows
- Frowning
- Symmetric vs. one-sided
- Frequency
- Duration
- Gaze shift
- Blinking
- Frequency
- Duration
- Frequency and duration of expressions listed in
Static Visual Cues
78Wigginss Circumplex
79iFACE Personality
- Facial actions associated with Wiggins
dimensions (visual cues) - Trigger facial actions
- Speech energy (Impulse and Emphasis)
- Randomly selected from the associated set
- Random/periodic
- Blinking and expressions
- Speech content
- Research
80Test Animations
- Asian Female
- White Female
- Asian Male
- White Male
81Cues Emotions
82Cues Slow Actions
83Cues Fast Actions
84Sample Mixed-Actions
- Exhibitionist
- High Dominance, High Affiliation
- Shy
- Low Dominance, Low Affiliation
- Competitor
- High Dominance, Low Affiliation
- Helper
- Low Dominance, High Affiliation
85Mixed-Actions Details
- HD-HA
- Joy, Fast Turn, Fast Tilt, Fast Double-Brow Raise
- LD-LA
- Contempt, Slow Nod, Fast Avert Gaze, Down Head
Rest - HD-LA
- Anger, Pride, Fast Single-Brow Raise
- LD-HA
- Surprise, Fast Blink, Slow Double-Brow Raise,
Right Head Rest
86Mixed-Action Results
87iFACE Software Architecture
88Sample iFACE Applications
- MusicFace
- Music-driven facial choreograpy
- COMPS
- Collaborative online multimedia problem-solving
simulation - Problem-based learning for medical students
89MusicFace (DiPaola Arya, 2004)
- Affective communication remapping
- Extracting affective info from music and creating
facial actions based on it (Demo)
90COMPS (DiPaola, Arya et al. 2005, Demo)
91More iFACE Applications
- Storytelling Masks
- Evolving Faces
92Recent Projects
- Painterly animation
- Create facial (and non-facial) animation similar
to the style of a painter - Non-photorealistic rendering (sample)
- Interactive stories
- Story space elements with relations
- Virtual communities
- Rules of behaviour
- Characters with personality
93Motion Capture
94Native Dance
95Digital Pow-wow
- Recording the motion data
- Defining style parameters
- Establishing computational style models
- Developing a designer software
- Creating a web service for making and displaying
virtual shows