Information theoretic sensor data selection for active object recognition and state estimation

DMCA

Information theoretic sensor data selection for active object recognition and state estimation (2002)

Cached

Download Links

Other Repositories/Bibliography

DBLP

by Joachim Denzler , Christopher M. Brown

Venue:	in IEEE Transactions on PAMI
Citations:	120 - 17 self

BibTeX

@INPROCEEDINGS{Denzler02informationtheoretic,
    author = {Joachim Denzler and Christopher M. Brown},
    title = {Information theoretic sensor data selection for active object recognition and state estimation},
    booktitle = {in IEEE Transactions on PAMI},
    year = {2002}
}

OpenURL

Abstract

Abstract—We introduce a formalism for optimal sensor parameter selection for iterative state estimation in static systems. Our optimality criterion is the reduction of uncertainty in the state estimation process, rather than an estimator-specific metric (e.g., minimum mean squared estimate error). The claim is that state estimation becomes more reliable if the uncertainty and ambiguity in the estimation process can be reduced. We use Shannon’s information theory to select information-gathering actions that maximize mutual information, thus optimizing the information that the data conveys about the true state of the system. The technique explicitly takes into account the a priori probabilities governing the computation of the mutual information. Thus, a sequential decision process can be formed by treating the a priori probability at a certain time step in the decision process as the a posteriori probability of the previous time step. We demonstrate the benefits of our approach in an object recognition application using an active camera for sequential gaze control and viewpoint selection. We describe experiments with discrete and continuous density representations that suggest the effectiveness of the approach. Index Terms—Computer vision, active camera control, state estimation, information theory. æ 1

Keyphrases

state estimation information theoretic sensor data selection active object recognition mutual information priori probability minimum mean iterative state estimation optimality criterion shannon information theory sequential gaze control information-gathering action true state data conveys index term computer vision information theory active camera control viewpoint selection certain time step sequential decision process estimation process continuous density representation posteriori probability optimal sensor parameter selection state estimation process estimate error object recognition application decision process static system previous time step active camera

DMCA