Model selection for the extraction of movement primitives

Abstract

A wide range of blind source separation methods have been used in motor control research for the extraction of movement primitives from EMG and kinematic data. Popular examples are principal component analysis (PCA), independent component analysis (ICA), anechoic demixing, and the time-varying synergy model (d'Avella and Tresch, 2002). However, choosing the parameters of these models, or indeed choosing the type of model, is often done in a heuristic fashion, driven by result expectations as much as by the data. We propose an objective criterion which allows to select the model type, number of primitives and the temporal smoothness prior. Our approach is based on a Laplace approximation to the posterior distribution of the parameters of a given blind source separation model, re-formulated as a Bayesian generative model. We first validate our criterion on ground truth data, showing that it performs at least as good as traditional model selection criteria [Bayesian information criterion, BIC (Schwarz, 1978) and the Akaike Information Criterion (AIC) (Akaike, 1974)]. Then, we analyze human gait data, finding that an anechoic mixture model with a temporal smoothness constraint on the sources can best account for the data.

Perceptual integration of kinematic components in the recognition of emotional facial expressions

“…In addition we found that each component contributed significantly to all tested expressions. This small number of necessary primitives is confirmed by psychophysical experiments and by results derived by Bayesian model comparison (Endres, Chiovetto, & Giese, 2013). This finding implies that the efficient dimensionality of the space of dynamic facial expressions, especially for the ones that are frequently used, might be much smaller than what is suggested by the FACS.…”

A novel approach for the generation of complex humanoid walking sequences based on a combination of optimal control and learning of movement primitives

“…In an earlier paper [6], we argued that kinematic MPs could be grouped into temporal [7,8,9], spatial [10,11], and spatio-temporal primitives [2]. Temporal MPs are the ones which we use in this paper.…”

“…The smoothness provided by GPs facilitates interpolation to step parameter values to which the MP model has not been exposed to during training. Furthermore, we have shown previously [6] that such GP-MPs are good models for biological gait data, so we hypothesized that they might be suitable for humanoid robots, too. The graphical model of our approach is shown in Figure 5, using plate notation: we would like to generate K movements, each with possibly different parameters l k , e.g.…”

Untitled

“…Furthermore, the perception of emotion based on spatio-temporal MPs has been investigated by Roether et al (2009) andChiovetto et al (2018). In our study, we are interested in comparing different MP types in a unified Bayesian framework (Endres et al 2013) with respect to the perception of naturalness.…”

“…MPs refer to building blocks of complex movements, but there is little consensus on an exact definition. Consequently, many different types of MPs have been proposed in literature (Endres et al 2013). These types can be classified as spatial (Giszter et al 1992;Tresch et al 1999), temporal (Clever et al 2016;Endres et al 2013), spatio-temporal (d'Avella et al 2003dynamical MPs (Ijspeert et al 2013).…”