Synonyms
Definition
The Fisher-Rao metric is a particular Riemannian metric defined on a parameterized family of conditional probability density functions (pdfs). If two conditional pdfs are near to each other under the Fisher-Rao metric, then the square of the distance between them is approximately equal to twice the average value of the log likelihood ratio of the conditional pdfs. If the log likelihood ratio of the two pdfs is near to zero, then it is difficult to distinguish between them using only sampled values.
Background
Suppose that a parameterized family of conditional pdfs is given and it is required to find the parameter value corresponding to the conditional pdf that best fits a given set of data. It is useful to have a distance function defined on pairs of conditional pdfs, such that if a given conditional pdf is...
References
Amari S-I (1985) Differential-geometric methods in statistics. Lecture notes in statistics, vol 28. Springer, New York
Bauer M, Bruveris M, Michor PW (2016) Uniqueness of the Fisher-Rao metric on the space of smooth densities. Bull Lond Math Soc 48(3):499–506
Rao C (1945) Information and accuracy attainable in the estimation of statistical parameters. Bull Calcutta Math Soc 37:81–91
Fisher RA (1922) On the mathematical foundations of theoretical statistics. Philos Trans R Soc Lond Ser A 222:309–368
Maybank SJ (2004) Detection of image structures using fisher information and the Rao metric. IEEE Trans Pattern Anal Mach Intell 26(12):49–62
Peter A, Rangarajan A (2009) Information geometry for landmark shape analysis: unifying shape representation and deformation. IEEE Trans Pattern Anal Mach Intell 31(2):337–350
Mio W, Liu X (2006) Landmark representations of shapes and Fisher-Rao geometry. In: Proceedings of the IEEE international conference on image processing, Atlanta. IEEE, pp 2113–2116
Coelin SR, Hancock ER (2012) Computing gender difference using Fisher-Rao metric from facial surface normals. In: Brazilian symposium of computer graphic and image processing, pp 336–343
Author information
Authors and Affiliations
Corresponding author
Section Editor information
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this entry
Cite this entry
Maybank, S.J. (2020). Fisher-Rao Metric. In: Computer Vision. Springer, Cham. https://doi.org/10.1007/978-3-030-03243-2_657-1
Download citation
DOI: https://doi.org/10.1007/978-3-030-03243-2_657-1
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-03243-2
Online ISBN: 978-3-030-03243-2
eBook Packages: Springer Reference Computer SciencesReference Module Computer Science and Engineering