No abstract

We compared the perceptual validity of human avatar walking animations driven by six different representations of human movement using a graphics Turing test. All six representations are based on movement primitives (MPs), which are predictive models of full-body movement that differ in their complexity and prediction mechanism. Assuming that humans are experts at perceiving biological movement from noisy sensory signals, it follows that these percepts should be describable by a suitably constructed Bayesian ideal observer model. We build such models from MPs and investigate if the perceived naturalness of human animations are predictable from approximate Bayesian model scores of the MPs. We found that certain MP-based representations are capable of producing movements that are perceptually indistinguishable from natural movements. Furthermore, approximate Bayesian model scores of these representations can be used to predict perceived naturalness. In particular, we could show that movement dynamics are more important for perceived naturalness of human animations than single frame poses. This indicates that perception of human animations is highly sensitive to their temporal coherence. More generally, our results add evidence for a shared MP-representation of action and perception. Even though the motivation of our work is primarily drawn from neuroscience, we expect that our results will be applicable in virtual and augmented reality settings, when perceptually plausible human avatar movements are required.

No abstract

We describe a sparse, variational posterior approximation to the Coupled Gaussian Process Dynamical Model (CGPDM), which is a latent space coupled dynamical model in discrete time. The purpose of the approximation is threefold: first, to reduce training time of the model; second, to enable modular re-use of learned dynamics; and, third, to store these learned dynamics compactly. Our target applications here are human movement primitive (MP) models, where an MP is a reusable spatiotemporal component, or "module" of a human full-body movement. Besides re-usability of learned MPs, compactness is crucial, to allow for the storage of a large library of movements. We first derive the variational approximation, illustrate it on toy data, test its predictions against a range of other MP models and finally compare movements produced by the model against human perceptual expectations. We show that the variational CGPDM outperforms several other MP models on movement trajectory prediction. Furthermore, human observers find its movements nearly indistinguishable from replays of natural movement recordings for a very compact parameterization of the approximation.

We investigate the role of prediction in biological movement perception by comparing different representations of human movement in a virtual reality (VR) and online experiment. Predicting movement enables quick and appropriate action by both humans and artificial agents in many situations, e.g. when the interception of objects is important. We use different predictive movement primitive (MP) models to probe the visual system for the employed prediction mechanism. We hypothesize that MP-models, originally devised to address the degrees-of-freedom (DOF) problem in motor production, might be used for perception as well. In our study we consider object passing movements. Our paradigm is a predictive task, where participants need to discriminate movement continuations generated by MP models from the ground truth of the natural continuation. This experiment was conducted first in VR, and later on continued as online experiment. We found that results transfer from the controlled and immersive VR setting with movements rendered as realistic avatars to a simple and COVID-19 safe online setting with movements rendered as stick figures. In the online setting we further investigate the effect of different occlusion timings. We found that contact events during the movement might provide segmentation points that render the lead-in movement independent of the continuation and thereby make perceptual predictions much harder for subjects. We compare different MP-models by their capability to produce perceptually believable movement continuations and their usefulness to predict this perceptual naturalness. Our research might provide useful insight for application in computer animation, by showing how movements can be continued without violating the expectation of the user. Our results also contribute towards an efficient method of animating avatars by combining simple movements into complex movement sequences.