Object localization by Bayesian correlation
Josephine Sullivan, Andrew Blake, Michael Isard et al. · IEEE International Conference on Computer Vision (ICCV) · 1999
Maximisation of cross correlation is a commonly used principle for intensity based object localization that gives a single estimate of location. However, to facilitate sequential inference (e.g. over time or scale) and to allow the representation of ambiguity, it is desirable to represent an entire probability distribution for object location. Although the cross correlation itself (or some function of it) has sometimes been treated as a probability distribution, this is not generally justifiable. Bayesian correlation achieves a consistent probabilistic treatment by combining several developments. The first is the interpretation of correlation matching functions in probabilistic terms, as observation likelihoods. Second, probability distributions of filter bank responses are learned from training examples. Inescapably, response learning also demands statistical modelling of background intensities, and there are links here with image coding and Independent Component Analysis. Lastly, multi scale processing is achieved in a Bayesian context by means of a new algorithm, layered sampling, for which asymptotic properties are derived.
Cited by 117 works. doi:10.1109/iccv.1999.790391
Authors
References
- A theory for multiresolution signal decomposition: the wavelet representation
- Receptive fields and functional architecture of monkey striate cortex
- Advances in neural information processing systems 7
- Emergence of simple-cell receptive field properties by learning a sparse code for natural images
- Active Shape Models-Their Training and Application
- Hands: A Pattern Theoretic Study of Biological Shapes
- Filters, Random Fields and Maximum Entropy (FRAME): Towards a Unified Theory for Texture Modeling
- Contour tracking by stochastic propagation of conditional density
- Signal matching through scale space
- Tracking and recognizing rigid and non-rigid facial motions using local parametric models of image motion