Abstract
This paper introduces a new generalisation of scale-space and pyramids, which combines statistical modelling with a spatial representation. The representation uses the familiar concept of multiple resolutions, but applied to a Gaussian mixture representation of the image - hence the title MGMM. It is shown that MGMM can approximate any probability density and can adapt to smooth motions. After a brief presentation of the theory, it is shown how MGMM can be applied to the estimation of visual motion.
| Translated title of the contribution | Multiresolution Gaussian Mixture Models for Visual Motion Estimation |
|---|---|
| Original language | English |
| Title of host publication | Unknown |
| Editors | I. Pitas |
| Publisher | Institute of Electrical and Electronics Engineers (IEEE) |
| Pages | 921 - 924 |
| Number of pages | 3 |
| ISBN (Print) | 0780367278 |
| Publication status | Published - Oct 2001 |