geometry videos symposium on computer animation 2003

35
Geometry Videos Symposium on Computer Animation 2003 Hector M. Briceño Collaborators: Pedro V. Sander, Leonard McMillan, Steven Gortler, and Hugues Hoppe

Upload: shira

Post on 25-Feb-2016

41 views

Category:

Documents


0 download

DESCRIPTION

Geometry Videos Symposium on Computer Animation 2003. Hector M. Briceño Collaborators: Pedro V. Sander, Leonard McMillan, Steven Gortler, and Hugues Hoppe. Motivation. Many sources of 3D Animation data: Motion Capture Visual Hulls Physical Simulations Sensor Data Skilled Animators - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Geometry Videos Symposium on Computer Animation 2003

Geometry VideosSymposium on Computer Animation 2003

Hector M. BriceñoCollaborators: Pedro V. Sander, Leonard McMillan, Steven

Gortler, and Hugues Hoppe

Page 2: Geometry Videos Symposium on Computer Animation 2003

22

2 2

Motivation Many sources of 3D

Animation data: Motion Capture Visual Hulls Physical Simulations Sensor Data Skilled Animators

Wide variety of formats, data, and reconstruction schemes…

Page 3: Geometry Videos Symposium on Computer Animation 2003

33

3 3

Problem: Sharing 3D Animations

Render a Video of the animation Use the similar software and/or

hardware Use static mesh compression for

each frame DEMO DEMO

Page 4: Geometry Videos Symposium on Computer Animation 2003

44

4 4

Approach: By representing manifold 3D objects using

a global 2D parametrization (mapping) it is possible to use existing video techniques to represent 3D animations.

Page 5: Geometry Videos Symposium on Computer Animation 2003

55

5 5

Assumptions of Geometry Videos

One or more manifold surfaces Consistent connectivity through the

duration of the animation No changes in topology Can undergo arbitrary deformations

as well as rigid-body transformations

Page 6: Geometry Videos Symposium on Computer Animation 2003

66

6 6

Outline Related Work Geometry Images and Geometry

Videos Cuts Parametrization Compression

Exploiting Temporal Coherence Results Future Work and Conclusions

Page 7: Geometry Videos Symposium on Computer Animation 2003

77

7 7

Related Work: Mesh Compression

Maintaining connectivity: Topological Surgery

[Taubin98] Progressive Meshes

[Hoppe96] Spectral Compression

[Karni00] Re-parametrizing:

Semi-regular: Progressive Compression [Khodakovsky00]

Fully regular: Geometry Images: fully regular [Gu02]

Page 8: Geometry Videos Symposium on Computer Animation 2003

88

8 8

Related Work: Animated Meshes MPEG4, VRML Animated Meshes “Multi-Resolution Dynamic Meshes with Arbitrary

Deformations” [Shamir00] “Representing Animations by PCA” [Alexa00] “Compression of Time-dependent geometry”

[Lengyel99] “Dynapack” [Ibarria03]

Page 9: Geometry Videos Symposium on Computer Animation 2003

99

9 9

Related Work: Video MPEG

Spatial, Temporal, SNR Scalability, Motion Compensation, High Compression, VBR…

Other… Layered Coding L-DCT [Amir96] Multi-resolution Video [Finkelstein96]

LOD both time and space. NAIVE [Briceno99]

Graceful degradation, error resilience

Page 10: Geometry Videos Symposium on Computer Animation 2003

1010

10 10

Geometry Images Represents a manifold surface in 3D

space as an 2D array of 3D points. Works in 3 steps:

Cutting: maps 3D surfaces to manifold Parametrization

Maps 3D space -> 2D parameter space Rasterization and Compression

Page 11: Geometry Videos Symposium on Computer Animation 2003

1111

11 11

Parametrization Maps 3D manifold surface onto 2D

square Different criteria or metrics: Conformal,

Area-preserving, Geometric-Stretch

Page 12: Geometry Videos Symposium on Computer Animation 2003

1212

12 12

Rasterization/Compression Sample points of parametrization

obtain a 2D grid of triplets (x,y,z) Compress resulting “image”

DEMO

Page 13: Geometry Videos Symposium on Computer Animation 2003

1313

13 13

Cutting: Geometry Image Iteratively Cut and Reparametrize

Final

Page 14: Geometry Videos Symposium on Computer Animation 2003

1414

14 14

Animated Meshes: Approach

How do we cut, parametrize and compress considering a time-sequence of meshes?

Page 15: Geometry Videos Symposium on Computer Animation 2003

1515

15 15

Cutting: Animations Animation frames should have the

same cut and parametrization

No Correspondence

c

Different Cuts and Parametrization

Page 16: Geometry Videos Symposium on Computer Animation 2003

1616

16 16

Cuts, how to pick? Looking at single frame might miss

something?

Approach: find a global cut considering all frames.

Page 17: Geometry Videos Symposium on Computer Animation 2003

1717

17 17

Global Cut Cut from frame

2 misses spike on frame 1 and spikes on frame 3

Cut

2

Glo

bal C

ut

Frame 1 Frame 2 Frame 3

Page 18: Geometry Videos Symposium on Computer Animation 2003

1818

18 18

Global Cut: how it works Run the iterative algorithm on all frames at

the same time. Pick worst avg. face on all

parametrizations…

Fram

e 1

Fram

e 2

Page 19: Geometry Videos Symposium on Computer Animation 2003

1919

19 19

Parametrization: Animation Cut and parametrization has

to be fixed for all frames in order to use one texture for whole animation

We currently apply the global cut to the first frame and compute parametrization on that frame.

Page 20: Geometry Videos Symposium on Computer Animation 2003

2020

20 20

Compression Spatial Compression:

Wavelets: Can support multiple levels of detail…

Temporal Compression Predictive Coding similar to MPEG Use affine transformations for predictor

Page 21: Geometry Videos Symposium on Computer Animation 2003

2121

21 21

Encoder Architecture

Basic Delta Encoder Uses affine transformations

ReferenceFrame

InputFrame

Cut &Parametrize

Rasterize/Encode Diff

Transform

Decode

Page 22: Geometry Videos Symposium on Computer Animation 2003

2222

22 22

Transformations: Global Global Trans. form a good

approximation

Frame 2Frame 1 Transformed Frame 1

Page 23: Geometry Videos Symposium on Computer Animation 2003

2323

23 23

Transformations: Global con’t Global cannot capture well

deformations within the object

Frame 1

Frame 2

Predictor of Frame 2 from Frame 1

Page 24: Geometry Videos Symposium on Computer Animation 2003

2424

24 24

Transformations: Local Apply transformation on

chartsFrame 1

Frame 2

Predictor

Page 25: Geometry Videos Symposium on Computer Animation 2003

2525

25 25

Transformations: Local w/ Spread & Blend

Spread. Include neighbors in the computation of the transformation

Blend between patches.Target

PredictorNo blendNo spread

Predictorw/blendw/spread

d c

Page 26: Geometry Videos Symposium on Computer Animation 2003

2626

26 26

Results Comparing Geometry Images Comparison to PCA Predictive Coding: Transformations

Global Local

Timing/Performance Level of Detail

Page 27: Geometry Videos Symposium on Computer Animation 2003

2727

27 27

Comparing Geometry Images: Snake

Page 28: Geometry Videos Symposium on Computer Animation 2003

2828

28 28

Comparison to PCA

Page 29: Geometry Videos Symposium on Computer Animation 2003

2929

29 29

Transformations: Global vs. Local

Page 30: Geometry Videos Symposium on Computer Animation 2003

3030

30 30

Transformation Performance

DEMO

2bpv P

8bpv I Baseline

4bpv P

8bpv P8bpv B

s d

Page 31: Geometry Videos Symposium on Computer Animation 2003

3131

31 31

Performance Timings Finding Cut (one frame): 2-7 mins Finding Cut (100 frames): 3-5 hrs Parametrization: 2-6 mins Encoding: 2-3 fps @ 256x256 Encoding: 6-16 fps @ 64x64 Decoding: 10 fps @ 256x256 Decoding: 30-60 fps @ 64x64

Page 32: Geometry Videos Symposium on Computer Animation 2003

3232

32 32

Level of Detail

Page 33: Geometry Videos Symposium on Computer Animation 2003

3333

33 33

Future Work Video Compression Transformations Chartification Parametrization Non-manifold objects

Page 34: Geometry Videos Symposium on Computer Animation 2003

3434

34 34

Conclusions Geometry Video as way to encode

and represent 3D animations Can use many of the 2D Video

Techniques/Features Spatial/Temporal scalability Error resiliency

Many other features to be exploited, i.e. fast clipping and hardware implementation…

Page 35: Geometry Videos Symposium on Computer Animation 2003

3535

35 35

Acknowledgements Collaborators: Pedro Sander,

Leonard McMillan, Steven Gortler, Hughes Hoppe, and Gu Xianfen.

Animations: Matthias Mueller and Daniel VlasicQuestions

?