Research Theme: Example-Based Pose Estimation
In example-based pose estimation, the configuration or "pose" of an evolving object is sought given visual evidence, having to rely uniquely on a set of examples. We assume here that, in a training stage, a number of feature measurements is extracted from the available images, while an "oracle" provides us with the true object pose at each instant.

In this scenario, a sensible approach consists in learning maps from features to poses, using the information provided by the training set. In particular, multi-valued mappings linking feature values to set of training poses can be easily constructed. A probability measure on any feature space is then naturally mapped to a convex set of probabilities on the set of training poses, in a form of a "belief function". Given a test image, its feature measurements translate into a collection of belief functions on the set of training poses, which when combined yield there an entire family of probability distributions. From the latter, both a single, central pose estimate and a set of extremal estimates can be computed, together with a measure of how reliable the estimate is. Measuring the conflict among the inferred belief functions provides a way of tackling occlusions and imperfect localization, and flags the need to update the model.

We call this technique "Belief Modelling Regression".
Relevant papers:
  •   Fabio Cuzzolin
    Visions of a generalized probability theory
    Lambert Academic Publishing, September 2014
    Visions of a generalized probability theory
  •  Fabio Cuzzolin and Wenjuan Gong
    A belief-theoretical approach to example-based pose estimation
    submitted to the IEEE Transactions on Fuzzy Systems, July 2012; revised November 2012, second revision January 2014
    (I.F. 6.306)
  • Fabio Cuzzolin and Ruggero Frezza
    An evidential reasoning framework for object tracking
    Proceedings of SPIE - Photonics East 99 - Telemanipulator and Telepresence Technologies VI
    Vol. 3840, pp. 13-24
    Haynes Convention Center, Boston, MA, September 19-22, 1999
    ABSTRACT Abstract
  • Fabio Cuzzolin and Ruggero Frezza
    Evidential modeling for pose estimation
    Proceedings of the 4th International Symposium on Imprecise Probabilities and Their Applications (ISIPTA'05)
    Carnegie Mellon University, Pittsburgh, July 2005
    Presentation Poster
Lab Member(s): Wenjuan Gong, Fabio Cuzzolin