epi imagine intuitive modeling and animation for interactive graphics & narrative environment...
TRANSCRIPT
EPI IMAGINEIntuitive Modeling and Animation
for Interactive Graphics & Narrative Environment
Visite direction 2013
A Computer Graphics group 3D Modeling, Animation, Cinematography
A pluridiscilinary teamMarie-Paule Cani (Prof computer science Ensimag/ INPG)François Faure (Prof computer science UJF)Stefanie Hahmann (Prof applied-math, Ensimag / INPG)Jean-Claude Léon (Prof mechanics, Ense3/ INPG)Olivier Palombi (Prof in anatomy ,UJF)Damien Rohmer (MdC CPE Lyon)Rémi Ronfard (CR1 INRIA)
14 PhD students
5 research engineers + 1 computer artist
Administrative assistants : Catherine Bessière (INRIA), Florence Polge (ERC)
Intuitive Modeling and Animation for Interactive Graphics & Narrative Environments
Major Issue : demand for more and more complex scenesArmies of Computer Artists during several years?Or automatic creation … with little control ??
Scientific focus: Creation tools for animated virtual worldsApplications to films & games, engineering, other sciences, general public
Creation of Virtual Worlds
Practice in the film industry
Best research results quickly available but…Artists do as much as they can WITHOUT computers
Shape, motion, stories are drafted on paper, clay, etc.
Digital stagesDesign stages
[Pixar, « Geri’s game »]
Creation of Virtual Worlds
Practice in the industry
Standard software
• Direct editing of DoF(NURBS, subdivision)
• Years of training
Spoils creativity!
Grand challenge?
“Make tools as transparent to the artists as special effects were made transparent to the public!”
[Rob Cook, Technical director Pixar, 2009]
Making tools transparent ?
Methodology
User-centered perspective on shapes & motionHigh-level models embedding a priori knowledgeGenerate details from minimal user inputAdvanced transfer and editing tools
Creating interactive virtual prototypesDesigning & experimenting in the same systemNatural interaction gestures Real-time response
Long term Vision
“Magic pen” to seamlessly create shapes, motions & stories
Shape
Story
Motion
Applications
Interactive virtual prototyping, various domains
Civil & mechanical engineering
Natural sciencesVirtual anatomy: ontology, 3D modeling & animationPlants: high level representations for plants geometry
Art, communication & education Interactive tools for educationDesign tools for artist and the publicTheater: virtual staging & rehearsals
Films & games Real-time, plausible clothing Virtual cinematography & film editing
Four scientific axes
1. High-level models for shapes In charge: Stefanie Hahmann
2. Models for real-time motion synthesis In charge: François Faure
3. Knowledge-based models for narrative design In charge: Rémi Ronfard
4. Creating and interacting with virtual prototypesIn charge: Jean-Claude Léon
New Results 2012
First image of a flat torus [@Rohmer]
The Visual Computer:
• FFD for GPU [Hahmann, Bonneau, Barbier, Elber, Hagen 2012]• Procedural Generation of Villages on Arbitrary Terrains
[Emilien, Bernhardt, Peytavie, Cani, Galin 2012]
Shapes 2012-2013Design Preserving Garment Transfer
[Brouet Sheffer Boissieux CaniSIGGRAPH 2012]
Shapes 2012-2013 (ANR Romma)Computing interfaces in assemblies
[Jourdes, Bonneau, Hahamnn, Faure, Léon, SPM/CAD 2013]
11
Shapes 2012-2013Analyse: morphologie et graphe de construction
Analyse morphologique pour l’idéalisation d’objets [F. Boussuge, J-C Léon, S. Hahmann, L. Fine, Idealized models for FEA derived fromgenerative modeling processes based on extrusion primitives, soumission Meshing RoundTable]
Shapes 2012-2013Analyse: morphologie et graphe de construction
Construction non triviale d’objets à partir de primitives d’extrusion. (F. Boussuge, J-C Léon, S. Hahmann, L. Fine, Extraction of generative processes from B-Rep shapes and application to idealization transformations, acceptation conditionnelle SPM 13)
Animation 2012-2013Adaptive Image-based Intersection Volume
Adaptive Image-based Intersection Volume [B. WANG, F. FAURE, D. K. PAI]
SIGGRAPH 2012
Animation 2012-2013Implicit skinning
Real-Time skinning with Contact Modeling[Vaillant, Barthe, Guennebaud, Cani, Rohmer,
Wyvill, Gourmel, Paulin]
SIGGRAPH 2013
Implicit skinning
Narrative design 2012-2013 Detecting and naming actors in movies
View-independent model of actor’s appearances (color) from examples• Compared to generic detectors, 20 % better in recall, same precision• Also useful for live performances [Gandhi Ronfard CVPR 2013]
Narrative design 2012-2013 The Prose Storyboard Language
A Tool for Annotating and Directing Movies.
• Pseudo-natural language description of movie shots
• for controled annotation of real movies & intuitive camera control in 3D animation.
[Intelligent Cinematography and Editing W, Ronfard, Gandhi Boiron 2013]
Creation & Interaction 2012-2013Understanding Table-top Interaction
Understanding Hand Degrees of Freedom and Natural Gestures for 3D Interaction on Tabletop
[Brouet, Blanch, Cani, Interact 2013]
Hand Navigator 2012-2013 Partnership Inria – Haption
Nouveau périphérique haptique pour la saisie réaliste d’objets
Creation & Interaction 2013Mutable elastic models for sculpting structured shapes
Clay-like scultping!
[Milliez, Wand, Cani,EUROGRAPHICS 2013]
Siggraph Asia 2013
Line of action
Expressive character posing
[Guay, Cani, Ronfard, Siggraph Asia 2013]
Siggraph Asia 2013
Anatomy Transfer
[Dicko, Faure, Gilles, Palombi, Kavan, Cani
Siggraph Asia 2013]
Soumissions TOG 2013
Sketching folds
[Jung, Hahmann, Rohmer, Begault,
Boissieux, Cani, submitted for
publication 2013]