CS-184: Computer Graphics

Transcription

1 CS-184: Computer Graphics Lecture #19: Motion Capture!!! Prof. James O Brien! University of California, Berkeley!! V2015-S Today 1 18-MoCap.key - April 8, 2015 Motion Capture MoCap.key - April 8, 2015

2 Motion Capture Record Use motion from physical objects! motion to animate virtual objects Simplified Pipeline: Setup and calibrate equipment Record performance Process motion data Generate animation MoCap.key - April 8, 2015 Basic Pipeline Setup Record Process Animatio n From Rose, et al., MoCap.key - April 8, 2015

3 What types of objects? Human, whole body! Portions of body! Facial animation! Animals! Puppets! Other objects MoCap.key - April 8, 2015 Capture Equipment Passive Optical! Reflective markers! IR (typically) illumination! Special cameras! Fast, high res., filters! Triangulate for positions Images from Motion Analysis MoCap.key - April 8, 2015

4 Capture Equipment Passive Optical Advantages! Accurate! May use many markers! No cables! High frequency! Disadvantages! Requires lots of processing! Expensive systems! Occlusions! Marker swap! Lighting / camera limitations 7 Capture Equipment Active Optical! Similar to passive but uses LEDs! Blink IDs, no marker swap! Number of markers trades off w/ frame rate 7 18-MoCap.key - April 8, 2015 Phoenix Technology Phase Space MoCap.key - April 8, 2015

5 Capture Equipment Magnetic Trackers! Transmitter emits field! Trackers sense field! Trackers report position and orientation May be wireless Control 9 Capture Equipment 9 18-MoCap.key - April 8, 2015 Electromagnetic Advantages! 6 DOF data! No occlusions! Less post processing! Cheaper than optical! Disadvantages! Cables! Problems with metal objects! Low(er) frequency! Limited range! Limited number of trackers MoCap.key - April 8, 2015

6 Capture Equipment Electromechanical Analogus 11 Capture Equipment Puppets MoCap.key - April 8, 2015 Digital Image Design MoCap.key - April 8, 2015

7 Performance Capture Many studios regard Motion Capture as evil! Synonymous with low quality motion! No directive / creative control! Cheap! Performance Capture is different! Use mocap device as an expressive input device! Similar to digital music and MIDI keyboards MoCap.key - April 8, 2015 Manipulating Motion Data Basic tasks! Adjusting! Blending! Transitioning! Retargeting! Building graphs MoCap.key - April 8, 2015

8 Nature of Motion Data Witkin and Popovic, 1995 Subset of motion curves from captured walking motion MoCap.key - April 8, 2015 Adjusting IK on single frames will not work Gleicher, SIGGRAPH MoCap.key - April 8, 2015

9 Adjusting Define desired motion function in parts Adjustment Inital sampled data Result after adjustment MoCap.key - April 8, 2015 Adjusting Select adjustment function from some nice space! Example C2 B-splines! Spread modification over reasonable period of time! User selects support radius MoCap.key - April 8, 2015

10 Adjusting IK uses control points of the B- spline now!! Example:! position racket! fix right foot! fix left toes! balance Witkin and Popovic SIGGRAPH MoCap.key - April 8, 2015 Adjusting Witkin and Popovic SIGGRAPH 95 What if adjustment periods overlap? MoCap.key - April 8, 2015

11 Blending Given two motions make a motion that combines qualities of both!!! Assume same DOFs! Assume same parameter mappings 21 Blending Consider blending slow-walk and fast-walk MoCap.key - April 8, 2015 Bruderlin and Williams, SIGGRAPH MoCap.key - April 8, 2015

12 Blending Define timewarp functions to align features in motion Normalized time is w MoCap.key - April 8, 2015 Blending Blend in normalized time!!! Blend playback rate MoCap.key - April 8, 2015

13 Blending Blending may still break features in original motions Touchdown for Run Touchdown for Walk Blend misses ground and floats MoCap.key - April 8, 2015 Blending Add explicit constrains to key points! Enforce with IK over time Touchdown for Run Touchdown for Walk MoCap.key - April 8, 2015

14 Blending / Adjustment Short edits will tend to look acceptable! Longer ones will often exhibit problems! Optimize to improve blends / adjustments! Add quality metric on adjustment! Minimize accelerations / torques! Explicit smoothness constraints! Other criteria MoCap.key - April 8, 2015 Multivariate Blending Extend blending to multivariate interpolation Speed Happiness MoCap.key - April 8, 2015

15 Multivariate Blending Extend blending to multivariate interpolation Speed If we have other examples place them in the space also Happiness Use standard scattered-data interpolation methods MoCap.key - April 8, 2015 Transitions Transition from one motion to another Perform blend in overlap region MoCap.key - April 8, 2015

16 Cyclification Special case of transitioning! Both motions are the same! Need to modify beginning and end of a motion simultaneously 31 Transition Graphs MoCap.key - April 8, 2015 Run Flip Sit Walk Trip Stand Dance MoCap.key - April 8, 2015

17 Motion Graphs Hand build motion graphs often used in games! Significant amount of work required! Limited transitions by design! Motion graphs can also be built automatically Run Flip Sit Walk Trip Stand Dance MoCap.key - April 8, 2015 Motion Graphs Similarity metric! Measurement of how similar two frames of motion are! Based on joint angles or point positions! Must include some measure of velocity! Ideally independent of capture setup and skeleton! Capture a large database of motions MoCap.key - April 8, 2015

18 Motion Graphs Random walks! Start in some part of the graph and randomly make transitions! Avoid dead ends! Useful for idling behaviors! Transitions! Use blending algorithm Domain of smoothing Discontinuity Magnitu Smoothed Signal Smoothing Function 35 Motion graphs Match imposed requirements! 35 Figure 4: In the synthesized motion, discontinuities in are inevitable. We deal with these discontinuities usin 18-MoCap.keyAt - April localized smoothing. the8, 2015 top left, a discontinuous signal, with its discontinuity shown at the top right. W struct an interpolant to this discontinuity, shown on the and add it back to the original signal to get the continu shown on the bottom left. Typically, discontinuities in are sufficiently small that no more complex strategy is 1. Replace a sequence by selecting two edges ei and 0 j n i, deleting all the edges between path and connecting the unconnected pieces of ing one or two edges in the top level graph G ( Since in the summarized graph, there are rela edges, we can quickly find edges that connect connected nodes by checking all the edges that tomotion(ei ), and enumerating all the edges t f rommotion(ei+ j ) and generate a valid path. N enumerate only 0 or 1 hop edges (1 edge or 2 e tions respectively). Start at a particular location! End at a particular location! Pass through particular pose! 2. Demoting two edges to their children and rep with one of their children if they can generate a Can be solved using dynamic programing! Doing this mutation on two edges simultaneous to compensate for the errors that would happen if Efficiency issues may require approximate solution! them was demoted. Notion of goodness of a solution We check every possible mutation, evaluate them and few. Since the summary has significantly fewer edges t inal graph, this step is not very expensive. If a motion se not generate a mutation whose score is lower that itsel that the current path is a local minimum in the valid pa record it as a potential motion. This way, we can obt motions that satisfy the same set of constraints MoCap.key April 8, 2015 Creating and -Smoothing the Final Path We create the final motion by taking the fram

19 Typical Motion Graph Walking #1 Running Idle Fall down Walking #2 Punches Finite number of states Cloth is hysteretic Recorded Time Naïve Precomputation MoCap.key - April 8, MoCap.key - April 8,

20 Graph Unrolling Graph Unrolling MoCap.key - April 8, MoCap.key - April 8,

21 Graph Unrolling 5000 hours compute 100K frames 330 GB MoCap.key - April 8, 2015 Graph Unrolling 5000 hours compute 100K frames 330 GB MoCap.key - April 8, 2015

22 Precomputed Cloth 72 MB Compressed Laptop 60 fps Low CPU load Wrong inset due to time constraints.! Really it works. Trust me! Precomputed Cloth MoCap.key - April 8, MoCap.key - April 8,

23 Precomputed Simulation No significant CPU load at runtime! Decouples quality from runtime cost! No new data at runtime! Simulation can t crash application! All motion can be inspected/edited! Allows QA and art direction of simulations! Extend to other types of simulation?! Dynamic variations? Suggested Reading MoCap.key - April 8, Fourier principles for emotion-based human figure animation, Unuma, Anjyo, and Takeuchi, SIGGRAPH 95! Motion signal processing, Bruderlin and Williams, SIGGRAPH 95! Motion warping, Witkin and Popovic, SIGGRAPH 95! Efficient generation of motion transitions using spacetime constrains, Rose et al., SIGGRAPH 96! Retargeting motion to new characters, Gleicher, SIGGRAPH 98! Verbs and adverbs: Multidimensional motion interpolation, Rose, Cohen, and Bodenheimer, IEEE: Computer Graphics and Applications, v. 18, no. 5, MoCap.key - April 8, 2015

24 Suggested Reading Retargeting motion to new characters, Gleicher, SIGGRAPH 98! Footskate Cleanup for Motion Capture Editing, Kovar, Schreiner, and Gleicher, SCA 2002.! Interactive Motion Generation from Examples, Arikan and Forsyth, SIGGRAPH 2002.! Motion Synthesis from Annotations, Arikan, Forsyth, and O'Brien, SIGGRAPH 2003.! Pushing People Around, Arikan, Forsyth, and O'Brien, unpublished.! Automatic Joint Parameter Estimation from Magnetic Motion Capture Data, O'Brien, Bodenheimer, Brostow, and Hodgins, GI 2000.! Skeletal Parameter Estimation from Optical Motion Capture Data, Kirk, O'Brien, and Forsyth, CVPR 2005.! Perception of Human Motion with Different Geometric Models, Hodgins, O'Brien, and Tumblin, IEEE: TVCG MoCap.key - April 8, 2015