Real-time control of three-dimensional avatars is an important problem in the context computer games and virtual environments. Avatar animation difficult, however, because a large repertoire avatar behaviors must be made available, user able to select from this set behaviors, possibly with low-dimensional input device. One appealing approach obtaining rich collect extended, unlabeled sequence motion data appropriate application. In paper, we show that such database can preprocessed for...

This paper describes algorithms for the animation of men and women performing three dynamic athletic behaviors: running, bicycling, vaulting. We animate these behaviors using control that cause a physically realistic model to perform desired maneuver. For example, allow simulated humans maintain balance while moving their arms, run or bicycle at variety speeds, handspring vault. Algorithms group number bicyclists ride as avoiding simple patterns obstacles. add secondary motion animations...

Article Design galleries: a general approach to setting parameters for computer graphics and animation Share on Authors: J. Marks MERL - A Mitsubishi Electric Research Laboratory, 201 Broadway, Cambridge, MA MAView Profile , B. Andalman Harvard Univ. Univ.View P. A. Beardsley W. Freeman S. Gibson Hodgins Georgia Tech. Tech.View T. Kang CMU CMUView Mirtich H. Pfister Ruml K. Ryall Seims of Washington WashingtonView Shieber Authors Info & Claims SIGGRAPH '97: Proceedings the 24th annual...

Real-time control of three-dimensional avatars is an important problem in the context computer games and virtual environments. Avatar animation difficult, however, because a large repertoire avatar behaviors must be made available, user able to select from this set behaviors, possibly with low-dimensional input device. One appealing approach obtaining rich collect extended, unlabeled sequence motion data appropriate application. In paper, we show that such database can preprocessed for...

Optimization is an appealing way to compute the motion of animated character because it allows user specify desired in a sparse, intuitive way. The difficulty solving this problem for complex characters such as humans due part high dimensionality search space. artifact representation most dynamic human behaviors are intrinsically low dimensional with, example, legs and arms operating coordinated We describe method that exploits observation create optimization easier solve. Our utilizes...

Despite the recent achievements in stable dynamic walking for many humanoid robots, relatively little navigation autonomy has been achieved. In particular, ability to autonomously select foot placement positions avoid obstacles while is an important step towards improved humanoids. We present a footstep planner Honda ASIMO robot that plans sequence of navigate toward goal location avoiding obstacles. The possible future are dependent on current state robot. Using finite set state-dependent...

Temporal segmentation of human motion into plausible primitives is central to understanding and building computational models motion. Several issues contribute the challenge discovering primitives: exponential nature all possible movement combinations, variability in temporal scale actions, complexity representing articulated We pose problem learning as one clustering, derive an unsupervised hierarchical bottom-up framework called aligned cluster analysis (HACA). HACA finds a partition given...

This paper introduces an approach to performance animation that employs video cameras and a small set of retro-reflective markers create low-cost, easy-to-use system might someday be practical for home use. The low-dimensional control signals from the user's are supplemented by database pre-recorded human motion. At run time, automatically learns series local models motion capture examples close match marker locations captured cameras. These then used reconstruct user as full-body animation....

Engaging storytelling is a necessary skill for humanoid robots if they are to be used in education and entertainment applications. Storytelling requires that the robot aware of its audience able direct gaze natural way. In this paper, we explore how human can modeled implemented on create natural, human-like behavior storytelling. Our model integrates data collected from storyteller discourse structure developed by Cassell her colleagues conversational agents (1994). We robot, Honda's ASIMO,...

We introduce a simple and effective deep learning approach to automatically generate natural looking speech animation that synchronizes input speech. Our uses sliding window predictor learns arbitrary nonlinear mappings from phoneme label sequences mouth movements in way accurately captures motion visual coarticulation effects. enjoys several attractive properties: it runs real-time, requires minimal parameter tuning, generalizes well novel sequences, is easily edited create stylized...

This paper presents a method for the detection and recognition of social interactions in day-long first-person video u event, like trip to an amusement park. The location orientation faces are estimated used compute line sight each face. context provided by all frame is convert lines into locations space which individuals attend. Further, assigned roles based on their patterns attention. rotes analyzed over time detect recognize types interactions. In addition face attention, head movements...

Using the pre-recorded human motion and trajectory tracking, we can control of a humanoid robot for free-space, upper body gestures. However, number degrees freedom, range joint motion, achievable velocities today's robots are far more limited than those average subject. In this paper, explore set techniques limiting gestures to that by Sarcos located at ATR. We assess quality results comparing actor robot, both visually quantitatively.

This paper is about the use of control algorithms to animate dynamic legged locomotion. Control could free animator from specifying details joint and limb motion while producing both physically realistic natural looking results. We implemented computer animations a biped robot, quadruped kangaroo. Each creature was modeled as linked set rigid bodies with compliant actuators at its joints. regulated running speed, organized legs, maintained balance. All motions were generated by numerically...

Article Free Access Share on Graphical modeling and animation of brittle fracture Authors: James F. O'Brien GVU Center College Computing, Georgia Institute Technology, Atlanta, GA GAView Profile , Jessica K. Hodgins Atlanata, Authors Info & Claims SIGGRAPH '99: Proceedings the 26th annual conference Computer graphics interactive techniquesJuly 1999Pages 137–146https://doi.org/10.1145/311535.311550Published:01 July 1999Publication History 319citation3,148DownloadsMetricsTotal...

Controllable, reactive human motion is essential in many video games and training environments. Characters these applications often perform tasks based on modified data, but response to unpredicted events also important order maintain realism. We approach the problem of synthesis for interactive, humanlike characters by combining dynamic simulation capture data. Our control systems use trajectory tracking follow data a balance controller keep character upright while modifying sequences from...

The task of controlling step length in the context a dynamic biped robot that actively balances itself as it runs is discussed. Three methods for length, each which adjusts different parameter running cycle, are adjusted parameters forward speed, height, and duration ground contact. All three control successful manipulating laboratory experiments, but method speed provided widest range lengths with accurate length. manipulated dynamics system so feet could be placed on available footholds...

Article Free Access Share on Adapting simulated behaviors for new characters Authors: Jessica K. Hodgins College of Computing and Graphics, Visualization Usability Center, Georgia Institute Technology, Atlanta, GA GAView Profile , Nancy S. Pollard Authors Info & Claims SIGGRAPH '97: Proceedings the 24th annual conference Computer graphics interactive techniquesAugust 1997 Pages 153–162https://doi.org/10.1145/258734.258822Published:03 August 1997Publication History...

Many compelling applications would become feasible if novice users had the ability to synthesize high quality human motion based only on a simple sketch and few easily specified constraints. We approach this problem by representing desired as an interpolation of two time-scaled paths through graph. The graph is constructed support pruned for efficient search. use anytime version A* search find globally optimal solution in that satisfies user's specification. Our retains natural transitions...

Even such simple tasks as placing a box on shelf are difficult to animate, because the animator must carefully position character satisfy geometric and balance constraints while creating motion perform task with natural-looking style. In this paper, we explore an approach for animating characters manipulating objects that combines power of path planning domain knowledge inherent in data-driven, constraint-based inverse kinematics. A planner is used find object corresponding poses geometric,...

High contrast images are common in night scenes and other that include dark shadows bright light sources. These difficult to display because their contrasts greatly exceed the range of most devices for images. As a result, image constrasts compressed or truncated, obscuring subtle textures details. Humans view understand high easily, “adapting” visual response avoid compression truncation with no apparent loss detail. By imitating some these adaptation processes, we developed methods...

During dynamic activities, the surface of human body moves in many subtle but visually significant ways: bending, bulging, jiggling, and stretching. We present a technique for capturing animating those motions using commercial motion capture system approximately 350 markers. Although number markers is significantly larger than that used conventional capture, it only sparse representation true shape body. supplement this sample with detailed, actor-specific model. The skin can then be...

We present a wireless virtual reality system and prototype full body Tai Chi training application. Our primary contribution is the creation of that tracks in working volume 4 meters by 5 2.3 high to produce an animated representation user with 42 degrees freedom. This - combined lightweight (<3 pounds) belt-worn video receiver head-mounted display provides wide area, untethered environment allows exploration new application areas. secondary our attempt show interface techniques made possible...