Motion Blending (Multidimensional Interpolation) Jehee Lee.

Slides:



Advertisements
Similar presentations
QR Code Recognition Based On Image Processing
Advertisements

Active Contours, Level Sets, and Image Segmentation
Lecture 10 Curves and Surfaces I
Interpolation By Radial Basis Function ( RBF ) By: Reihane Khajepiri, Narges Gorji Supervisor: Dr.Rabiei 1.
Animation Following “Advanced Animation and Rendering Techniques” (chapter 15+16) By Agata Przybyszewska.
Overview June 9- B-Spline Curves June 16- NURBS Curves June 30- B-Spline Surfaces.
CS 445/645 Fall 2001 Hermite and Bézier Splines. Specifying Curves Control Points –A set of points that influence the curve’s shape Knots –Control points.
Jehee Lee Seoul National University
Interpolation to Data Points Lizheng Lu Oct. 24, 2007.
Computer Animation Algorithms and Techniques
Dr. S.M. Malaek Assistant: M. Younesi
Presentation and Analysis of a Multi-Dimensional Interpolation Function for Non-Uniform Data: Microsphere Projection Presented By: Will Dudziak.
1 Curves and Surfaces. 2 Representation of Curves & Surfaces Polygon Meshes Parametric Cubic Curves Parametric Bi-Cubic Surfaces Quadric Surfaces Specialized.
1cs542g-term Notes  Added required reading to web (numerical disasters)
Introduction to Data-driven Animation Jinxiang Chai Computer Science and Engineering Texas A&M University.
Cornell CS465 Fall 2004 Lecture 16© 2004 Steve Marschner 1 Curved surfaces CS 465 Lecture 16.
CSCE 441 Computer Graphics: Keyframe Animation/Smooth Curves Jinxiang Chai.
Modelling. Outline  Modelling methods  Editing models – adding detail  Polygonal models  Representing curves  Patched surfaces.
Geometric Modeling Surfaces Mortenson Chapter 6 and Angel Chapter 9.
Curves Mortenson Chapter 2-5 and Angel Chapter 9
Bezier and Spline Curves and Surfaces Ed Angel Professor of Computer Science, Electrical and Computer Engineering, and Media Arts University of New Mexico.
Modelling: Curves Week 11, Wed Mar 23
University of British Columbia CPSC 414 Computer Graphics © Tamara Munzner 1 Curves Week 13, Mon 24 Nov 2003.
Bezier and Spline Curves and Surfaces CS4395: Computer Graphics 1 Mohan Sridharan Based on slides created by Edward Angel.
COEN Computer Graphics I
Human Computer Interaction 7. Advanced User Interfaces (I) Data Scattering and RBF Course no. ILE5013 National Chiao Tung Univ, Taiwan By: I-Chen Lin,
Computer Graphics Lecture 13 Curves and Surfaces I.
Curve Modeling Bézier Curves
Bresenham’s Algorithm. Line Drawing Reference: Edward Angel’s book: –6 th Ed. Sections 6.8 and 6.9 Assuming: –Clipped (to fall within the window) –2D.
11/19/02 (c) 2002, University of Wisconsin, CS 559 Last Time Many, many modeling techniques –Polygon meshes –Parametric instancing –Hierarchical modeling.
CIS V/EE894R/ME894V A Case Study in Computational Science & Engineering HW 5 Repeat the HW associated with the FD LBI except that you will now use.
A D V A N C E D C O M P U T E R G R A P H I C S CMSC 635 January 15, 2013 Spline curves 1/23 Curves and Surfaces.
Digital Multimedia, 2nd edition Nigel Chapman & Jenny Chapman Chapter 8 This presentation © 2004, MacAvon Media Productions Animation.
This Week Week Topic Week 1 Week 2 Week 3 Week 4 Week 5
V. Space Curves Types of curves Explicit Implicit Parametric.
Introduction to Computer Graphics with WebGL
CSC418 Computer Graphics n Animation Principles n Keyframe Animation.
Numerical Computation
Computer Animation Rick Parent Computer Animation Algorithms and Techniques Interpolating Values.
 The creation of moving pictures one frame at a time Literally 'to bring to life' e.g. make a sequence of drawings on paper, in which a character's position.
Course 13 Curves and Surfaces. Course 13 Curves and Surface Surface Representation Representation Interpolation Approximation Surface Segmentation.
1 CS 430/536 Computer Graphics I Curve Drawing Algorithms Week 4, Lecture 8 David Breen, William Regli and Maxim Peysakhov Geometric and Intelligent Computing.
Constructing Bezier Curves on the Surface of a Sphere By Reza Ali Fundamentals of Spatial Computing UCSB MAT 594CM Spring 2009.
Image Morphing ( Computational Photography) Jehee Lee Seoul National University With a lot of slides stolen from Alexei Efros and Seungyong Lee.
CSCE 441 Computer Graphics: Keyframe Animation/Smooth Curves Jinxiang Chai.
CS 551/651 Advanced Graphics Arc Length. Assignment 1 Due Week from Thursday Building a Cubic Bézier curve OpenGL/Glut Insert up to 100 points Render.
Keyframing and Splines Jehee Lee Seoul National University.
CS 376 Introduction to Computer Graphics 04 / 25 / 2007 Instructor: Michael Eckmann.
CSCE 441 Computer Graphics: Keyframe Animation/Smooth Curves Jinxiang Chai.
11/6/ :55 Graphics II Introduction to Parametric Curves and Surfaces Session 2.
CS 551/651 Advanced Graphics Technical Background.
using Radial Basis Function Interpolation
CSCE 441: Keyframe Animation/Smooth Curves (Cont.) Jinxiang Chai.
CS559: Computer Graphics Final Review Li Zhang Spring 2010.
Keyframe Interpolation and Speed Control Jehee Lee Seoul National University.
Splines Sang Il Park Sejong University. Particle Motion A curve in 3-dimensional space World coordinates.
Ship Computer Aided Design
Basic Theory (for curve 01). 1.1 Points and Vectors  Real life methods for constructing curves and surfaces often start with points and vectors, which.
Reverse Engineering of Point Clouds to Obtain Trimmed NURBS Lavanya Sita Tekumalla Advisor: Prof. Elaine Cohen School of Computing University of Utah Masters.
Professor William H. Press, Department of Computer Science, the University of Texas at Austin1 Opinionated in Statistics by Bill Press Lessons #46 Interpolation.
Curves University of British Columbia CPSC 314 Computer Graphics Jan-Apr 2013 Tamara Munzner.
Processing Images and Video for An Impressionist Effect Automatic production of “painterly” animations from video clips. Extending existing algorithms.
Introduction to Parametric Curve and Surface Modeling.
CSCE 441 Computer Graphics: Keyframe Animation/Smooth Curves
CSCE 441 Computer Graphics: Keyframe Animation/Smooth Curves
© University of Wisconsin, CS559 Fall 2004
© University of Wisconsin, CS559 Spring 2004
Computer Graphics Lecture 15.
Introduction to Parametric Curve and Surface Modeling
Overview June 9- B-Spline Curves June 16- NURBS Curves
Presentation transcript:

Motion Blending (Multidimensional Interpolation) Jehee Lee

Data-Driven Approach It is difficult to understand basic principles of the real world –Instead, we would like to sample the real world ! What kind of data is available ? –Pictures (camera), video (camcorder) –Motion, facial expression (motion capture) –Geometry (3D scanner) –Voice, sound (recorder) –Tactile, physical properties, … Data-driven approaches try to reconstruct the real world in a computer from a rich set of samples

The real world is multi-dimensional Simple (object space) classification –One dimension: Sketch, curve –Two dimension: Pictures, surface, tactile –Three dimension: 3D geometry, video –Four dimension: Particle motion (position + time) –High dimension: Articulated figure motion Many interpretations (parameterizations) are possible independent of object space dimensions

Dimensionality of Motion Motion can have many parameters (dimensionality) –Physical dimensions: Velocity, turning angle, reach position –Ambiguous dimensions: Style, emotion, mood [Rose et al. 98]

The Space of Motion The space of every possible human walk –M-dimensional object space –N-dimensional parameter space Object space Parameter space

Multi-Dimensional Interpolation 3-parameters (velocity, turning angle, happiness) n-samples (motion clips) Interpolation gives a motion for any given parameters –We want to compute a continuous function

How many samples are needed ? One dimension: 1000 (points)

How many samples are needed ? Two dimension: 1000x1000 = (pixels)

How many samples are needed ? High-dimension –We cannot acquire enough samples –The interpolating function should be reconstructed from scattered (possibly sparse) samples We have to solve a multi-dimensional interpolation problem with scattered samples

Multi-Dimensional Functions One parameter (Curve) Two parameters (Surface) Three parameters (Volume) t u v u v w C(t) S(u,v) V(u,v,w)

Tensor-Product Surfaces Cubic polynomial curve Tensor-product surface –A curve is “multiplied” by a curve

Tensor-Product Functions Tensor-product is a standard technique for increasing the dimension of parametric functions –Bezier surfaces, B-spline surfaces, NURBS surfaces, … It works fine for low-dimensional parametric spaces –Great for surfaces –Maybe good enough for volumes It can be problematic for higher-dimensions –Too many control points –High degree of basis functions

Radial Basis Functions Radial refers to the pattern that you get when straight lines are drawn from the center of a circle to a number of points round the edge Radial basis function –A real valued function –having a center in a parameter space –The function value is determined by a distance from the center

Scattered Data Interpolation using RBF Find a function that interpolate given points such that RBF interpolation –We use M basis functions (same as # of given data points) –We don’t have a grid structure (compare to tensor-product surfaces) –Radial basis function is easily defined in any dimensional spaces

Popular Choices for RBF Thin-Plate Multiquadric Gaussian Biharmonic Triharmonic See demo: RBFwithoutLinearApproximation.exe RBFwithoutLinearApproximation.exe for some constant c

Augmented Polynomial Function A linear function or a low-order polynomial is augmented for better extrapolation Under-specified linear system is obtained Orthogonality conditions give a unique solution

Motion Blending S. I. Park, et al., On-line Locomotion Generation Based on Motion Blending, Symposium on Computer Animation, 2002 Movie

Application to Motion Blending The length of time –The first motion is 3 seconds long, the second motion is 5 seconds long, and so on. –Those motions are normalized in time and blended –What is the length of the blend ? The length of the blend can be negative –RBF interpolation doesn’t have convex hull property –Is it make sense to create a “backward walk” by blending “forward walks” ?

Application to Motion Blending Handling unit quaternions (or rotation matrices) –What if we apply RBF interpolation component-wisely ? Represent the surface as an affine combination –Construct cardinal basis functions using RBF

Application to Motion Blending Inverse problem –Precise control usually requires numerical iterations –May not have a unique solution –Eg) Parameter 3 m/s 5.5 m/s 7.5 m/s12 m/s Find a set of weights (coordinates in the parameter space) to create a walking motion of 8 m/s Actual walking speed

Other applications Surface reconstruction [Carr et al. 2002]

Other applications Shape morphing [Turk and O’Brien 99]

Summary Multidimensional interpolation –Sparse samples in high-dimensional spaces –Time-series data Scatter data interpolation using radial basis functions –The basis function is determined by applications –All basis function explained are not locally supported –For some applications, locally supported basis functions should be chosen