There is an error in our paper on intrinsic symmetry analysis with Mobius Transformations. Geodesic distances (used to measure geodesic error) are not normalized by square root of the area, thus the first row of Tables 1, 2 and 3 is completely meaningless (since all surfaces differ in scales). Please, contact Vladimir Kim (vk-AT-PRINCETON-EDU) if you have any further questions.

Mobius Transformations For Global Intrinsic Symmetry Analysis
V. G. Kim, Y. Lipman, X. Chen, T. Funkhouser
Computer Graphics Forum (Proc. of SGP 2010)