Making the Coupled Gaussian Process Dynamical Model Modular and Scalable with Variational Approximations

Velychko, Dmytro; Knopp, Benjamin; Endres, Dominik

doi:10.3390/e20100724

“…We perceptually validate 6 generative MP models: Temporal MPs, Dynamical MPs and 4 flavors of the Gaussian Process Dynamical Model (GPDM) (Velychko et al 2018;Wang et al 2008): GPDM, variational GPDM, coupled GPDM, and variational coupled GPDM.…”

Section: Movement Primitivesmentioning

confidence: 99%

“…Please refer to the cited papers for detailed information. Velychko et al (2018) also provide graphical model representations and summarize the features of the MP models presented in this chapter. (Clever et al 2016).…”

Section: Movement Primitivesmentioning

confidence: 99%

“…As with the MAP-trained GPDM introduced in the previous section, there is no explicit ideal observer model complexity parameter in the MAP-trained cGPDM. (Velychko et al 2018). The vCG-PDM compresses the movement representation of cGPDMs by introducing sparse variational approximations with a deterministic learning scheme.…”

Section: Coupled Gaussian Process Dynamical Model (Cgpdm)mentioning

confidence: 99%

“…Tables 1 and 2 summarize the tested models and ideal observer parameters. A more detailed description of the training procedure can be found in Velychko et al (2018). We trained each model until the training target (ELBO or training MSE) did not change within machine precision anymore, but at most for one day.…”

Section: Stimulus Generationmentioning

confidence: 99%

See 2 more Smart Citations

Untitled

2019

ACM Trans. Appl. Percept.

1

0

View full text Add to dashboard Cite

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.

show abstract

“…For the cases that ground truth data are unavailable or can only be determined approximately, Gaussian process latent variable models were developed in References [15,16] and they were extended to the setting of dynamical robotics systems in Reference [17]. In order to reduce the cubic complexity of Gaussian process training for the fixed number of training points, sparse Gaussian processes were developed (see e.g., [18][19][20][21][22][23]). K-optimality was used to improve the stability of the Gaussian process prediction in Reference [24].…”

Section: Introductionmentioning

confidence: 99%

Multisensor Estimation Fusion with Gaussian Process for Nonlinear Dynamic Systems

Liao

¹

,

Xie

²

,

Wang

³

et al. 2019

Entropy

View full text Add to dashboard Cite

The Gaussian process is gaining increasing importance in different areas such as signal processing, machine learning, robotics, control and aerospace and electronic systems, since it can represent unknown system functions by posterior probability. This paper investigates multisensor fusion in the setting of Gaussian process estimation for nonlinear dynamic systems. In order to overcome the difficulty caused by the unknown nonlinear system models, we associate the transition and measurement functions with the Gaussian process regression models, then the advantages of the non-parametric feature of the Gaussian process can be fully extracted for state estimation. Next, based on the Gaussian process filters, we propose two different fusion methods, centralized estimation fusion and distributed estimation fusion, to utilize the multisensor measurement information. Furthermore, the equivalence of the two proposed fusion methods is established by rigorous analysis. Finally, numerical examples for nonlinear target tracking systems demonstrate the equivalence and show that the multisensor estimation fusion performs better than the single sensor. Meanwhile, the proposed fusion methods outperform the convex combination method and the relaxed Chebyshev center covariance intersection fusion algorithm.

show abstract

Predicting Perceived Naturalness of Human Animations Based on Generative Movement Primitive Models

Knopp

¹

,

Velychko

²

,

Dreibrodt

³

et al. 2019

ACM Trans. Appl. Percept.

Self Cite

View full text Add to dashboard Cite

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.

show abstract

Making the Coupled Gaussian Process Dynamical Model Modular and Scalable with Variational Approximations

Cited by 6 publications

References 32 publications

Untitled

Untitled

Multisensor Estimation Fusion with Gaussian Process for Nonlinear Dynamic Systems

Predicting Perceived Naturalness of Human Animations Based on Generative Movement Primitive Models

Contact Info

Product

Resources

About