27/02/2015 motion autonomy IN CONSTRAINED ENVIRONMENTS Requires specific locomotion skills Motion planning and synthesis for arbitrary creatures in constrained environments Steve Tonneau Motion results from contact interactions Advisors: Franck Multon Julien Pettré Introduction Related work Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 2 Motion results from relevant contact sequence State of the art application: assassin’s creed 3 Introduction Related work Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 3 Introduction Related work Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 4 "Can you climb ON EVERY TREE?" MOTION CAPTURE: one motion, one session “It has to be a tree that splits into two trunks” (Beyond: 2 souls) New geometries? Non humanoid characters? Introduction Related work Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion Introduction Related work Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 5 6 1
27/02/2015 Motion Autonomy as a motion planning problem Motion Autonomy as a motion planning problem A virtual creature R The workspace W A virtual creature R The environment W Start and goal Start and goal postures postures A collision free A relevant contact motion sequence Introduction Related work Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 7 Introduction Related work Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 8 THe configuration space (Lozano-Peréz 83) Related work (Lozano-Peréz83) Robotics solutions to the motion planning problem Configuration space approach � � . .. . . . . . .. .. .. .. .. . . . .. .. .. � Sampling based motion planners . . . . . . � ∈ ℝ N q = . . . . . .. .. .. . .. .. . .. . . .. .. .. .. .. .. . . . � . . . . . . � z . . .. . .. .. . . . … .. .. .. Towards more natural looking motions x y Motion editing techniques Model based approaches Conf. space Conf. space motion planners motion planners motion editing motion editing motion analysis motion analysis Conf. space motion planners motion planners motion editing motion editing motion analysis motion analysis Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 9 Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 10 THe configuration space (Lozano-Peréz 83) THe configuration space (Lozano-Peréz 83) (Lozano-Peréz83) � ���� � � � ��� . .. . . . . . .. . . . .. .. .. .. .. .. .. � . . . . . . � ∈ ℝ N q = . . . . . . . . . . . . .. .. .. .. .. .. .. .. .. .. .. .. � . . . . . . � z � ������� . . . .. .. .. . .. . . .. .. … x y 3D environment � 3D environment � Configuration space � Configuration space � 3D environment � 3D environment � Configuration space � Configuration space � Conf. space motion planners motion planners motion editing motion editing motion analysis motion analysis Conf. space motion planners motion planners motion editing motion editing motion analysis motion analysis Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 11 12 2
27/02/2015 Sampling based motion planning Contact before motion approaches Probabilistic Road Maps (Kavrakiet al. 96) 1 step planning in the contact space Rapidly exploring Random Trees (Lavalle98) (Bretlet al. 03) (Escandeet al. 09) (Bouyarmaneet al. 09) x x PERFORMANce Quality of Motion Conf. space Conf. space motion planners motion editing motion editing motion analysis motion analysis Conf. space Conf. space motion planners motion editing motion editing motion analysis motion analysis Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 13 Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 14 TOWARDS More natural looking SOLUTIONS Combining Motion planning and Motion Editing Trajectory optimization (CIO) Sampling + Motion capture (Pettréet al. 03) (Choi et al. 03) (Huang et al. 11 ) (Mordatchet al. 12) Local minima PERFORMANce … Conf. space Conf. space motion planners motion editing motion editing motion analysis motion analysis Conf. space Conf. space motion planners motion planners motion editing motion analysis motion analysis Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 15 Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 16 Combining Motion planning and Motion Editing Motion editing LIMITATIONS Warping UNNATURAL DEFORMATIONS (Witkinet Popovic95) (Safonovaet Hodgins, 07) (Gleicher97), (Al-ashqaret al. 13) (Reitsmaet Pollard, 07) JohanssenUnity 3d Blending FIXED CONTACT SEQUENCEs (Boulic et al. 96 ????) w 1 + w 2 = (Kovaret Gleicher03) (Ménardaiset al. 04) Concatenating (Kovaret Gleicher02) (Shum al. 09) Conf. space Conf. space motion planners motion planners motion editing motion analysis motion analysis Conf. space Conf. space motion planners motion planners motion editing motion analysis motion analysis Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 17 18 3
27/02/2015 FROM MOTION ANALYSIS TO MODELS Related work SUMMARY Dynamic models and controllers “A trade off between naturalness and control” (Van Welbergenet al. 10) (Coroset al. 11) (Mordatchet al. 13) Unknown (Wampleret al. 14) Arbitrary Constrained Quality of motion Performance creatures environments Motion invariants for reaching / manipulating Motion editing + +++ +++ NO Model based ++ ++ +++ YES (Kallmannet al. 10) (Sreenivasaet al. 12) Sampling based +++ + + YES Sampling based +++ ? ? YES Conf. space Conf. space motion planners motion planners motion editing motion editing motion analysis Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 19 Introduction Introduction Related work Overview Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 20 Contribution 1: Contribution 1: A new Criterion for task efficiency A new Criterion for task efficiency Considering locomotion tasks in constrained environments EFORT EFORT How to generate a task efficient contact configuration? EFORT: The Extended FORce Transmission ratio Introduction Introduction Related work Related work Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 21 Introduction Introduction Related work Related work Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 22 Contribution 2: Contribution 2: A sampling based planner for EFORT A sampling based planner for EFORT How to compute relevant contact sequences for How to compute relevant contact sequences for locomotion tasks in constrained environments? locomotion tasks in constrained environments? RB-PRM: Reachability Based PRM ANIMATION ANIMATION RB-PRM RB-PRM EFORT EFORT EFORT Introduction Introduction Related work Related work Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion Introduction Introduction Related work Related work Overview EFORT EFORT RB-PRM RB-PRM Conclusion Conclusion 23 24 4
27/02/2015 EFORT THE IMPORTANCE OF FORCE EXERTION Which contacts to stand up? EFORT EFORT Task efficient contact configurations for Using task efficient contact configurations to arbitrary virtual creatures animate creatures in arbitrary environments Tonneau, Pettréet Multon, Tonneau, Pettréet Multon, Computer and Graphics journal, vol41, 2014 Graphics interface ’14 conference Introduction Introduction Related work Related work Overview Overview EFORT RB-PRM RB-PRM Conclusion Conclusion 25 Introduction Introduction Related work Related work Overview Overview EFORT RB-PRM RB-PRM Conclusion Conclusion 26 THE IMPORTANCE OF FORCE EXERTION EFORT: PROBLEM STATEMENT Given: Random stable configuration How to rapidly compute: More natural configuration More efficient configuration for force exertion Introduction Introduction Related work Related work Overview Overview EFORT RB-PRM RB-PRM Conclusion Conclusion 27 Introduction Introduction Related work Related work Overview Overview EFORT RB-PRM RB-PRM Conclusion Conclusion 28 THE Manipulability ellipsoid EFORT: THE EXTENDED force transmission ratio # $ � �, � � � � ! � ! � � �" � � � % (Chiu 87) 0 � Velocity and force ellipsoids (Yoshikawa 84) � &'()* �, � � +�* q, � �. . / Introduction Introduction Related work Related work Overview Overview EFORT RB-PRM RB-PRM Conclusion Conclusion Introduction Introduction Related work Related work Overview Overview EFORT RB-PRM RB-PRM Conclusion Conclusion 29 30 5
Recommend
More recommend
