The definition of Linear Symmetry-Based Disentanglement (LSBD) proposed by (Higgins et al., 2018) outlines the properties that should characterize a disentangled representation that captures the symmetries of data. However, it is not clear how to measure the degree to which a data representation fulfills these properties. We propose a metric for the evaluation of the level of LSBD that a data representation achieves. We provide a practical method to evaluate this metric and use it to evaluate the disentanglement of the data representations obtained for three datasets with underlying $SO(2)$ symmetries.
|Publication status||Published - 26 Nov 2020|
|Event||NeurIPS 2020 workshop on Differential Geometry meets Deep Learning - Online|
Duration: 11 Dec 2020 → 11 Dec 2020
|Workshop||NeurIPS 2020 workshop on Differential Geometry meets Deep Learning|
|Period||11/12/20 → 11/12/20|