← ArchivePaper2005
Geostatistical Motion Interpolation
Geostatistical approach to motion interpolation using kriging for smooth synthesis of human movement from sparse example motion clips.
Abstract
A common motion interpolation technique for realistic human animation is to blend similar motion samples with weighting functions whose parameters are embedded in an abstract space. Existing methods, however, are insensitive to statistical properties, such as correlations between motions. In addition, they lack the capability to quantitatively evaluate the reliability of synthesized motions. This paper proposes a method that treats motion interpolations as statistical predictions of missing data in an arbitrarily definable parametric space. A practical technique of geostatistics, called universal kriging, is then introduced for statistically estimating the correlations between the dissimilarity of motions and the distance in the parametric space. Our method statistically optimizes interpolation kernels for given parameters at each frame, using a pose distance metric to efficiently analyze the correlation. Motions are accurately predicted for the spatial constraints represented in the parametric space, and they therefore have few undesirable artifacts, if any. This property alleviates the problem of spatial inconsistencies, such as foot-sliding, that are associated with many existing methods. Moreover, numerical estimates for the reliability of predictions enable motions to be adaptively sampled.
How to read this
- Category
- Method: a statistical motion-interpolation technique
- Contributions
- Reframes motion interpolation as statistical prediction of missing data in a parametric space
- Introduces universal kriging to estimate correlations between motion dissimilarity and parametric distance, optimizing interpolation kernels per frame
- Reduces spatial-inconsistency artifacts such as foot-sliding and offers a notion of reliability for synthesized motion
- Context
- Advances parametric example-based motion interpolation (Rose et al., Verbs and Adverbs) by replacing fixed weighting functions with a geostatistical, correlation-aware estimator.Builds on: Verbs and Adverbs: Multidimensional Motion Interpolation
- Correctness
- Argued to reduce artifacts by accounting for inter-motion correlations and a pose distance metric; results depend on the chosen parametric space and pose metric, and kriging assumptions (a meaningful variogram from the examples) may not hold for sparse or poorly distributed clips.
- Clarity
- The motion-blending framing is accessible, but the kriging machinery is the harder part; a first pass gives the intuition, a second pass is needed for the variogram and kernel optimization.
- How to read it
- First pass for why correlation-aware interpolation beats fixed weights; a second pass on the universal-kriging formulation pays off if you work on motion synthesis or want the reliability estimate.
Builds on
Built upon by
Nothing yet.
Related work
- Verbs and Adverbs: Multidimensional Motion Interpolation 1998 / SIGGRAPH
- Automated Extraction and Parameterization of Motions in Large Data Sets 2004 / SIGGRAPH
- Near-Optimal Character Animation with Continuous Control 2007 / SIGGRAPH
- Neural State Machine for Character-Scene Interactions 2019 / SIGGRAPH Asia
Keywords
This page summarises the entry and links to its original source. The archive never hosts or redistributes the publication itself.Show it in the full archive list →