Journal Articles SIAM Journal on Imaging Sciences Year : 2017

Template estimation in computational anatomy: Fréchet means in top and quotient spaces are not consistent

Abstract

In this article, we study the consistency of the template estimation with the Fréchet mean in quotient spaces. The Fréchet mean in quotient spaces is often used when the observations are deformed or transformed by a group action. We show that in most cases this estimator is actually inconsistent. We exhibit a sufficient condition for this inconsistency, which amounts to the folding of the distribution of the noisy template when it is projected to the quotient space. This condition appears to be fulfilled as soon as the support of the noise is large enough. To quantify this inconsistency we provide lower and upper bounds of the bias as a function of the variability (the noise level). This shows that the consistency bias cannot be neglected when the variability increases.
Fichier principal
Vignette du fichier
InconsistencySIIMSDevilliers.pdf (634) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-01352707 , version 1 (09-08-2016)
hal-01352707 , version 2 (09-08-2016)
hal-01352707 , version 3 (11-08-2016)
hal-01352707 , version 4 (25-04-2017)

Identifiers

Cite

Loïc Devilliers, Stéphanie Allassonnière, Alain Trouvé, Xavier Pennec. Template estimation in computational anatomy: Fréchet means in top and quotient spaces are not consistent. SIAM Journal on Imaging Sciences, 2017, 10 (3), pp.1139-1169. ⟨10.1137/16M1083931⟩. ⟨hal-01352707v4⟩
1250 View
376 Download

Altmetric

Share

More