NeurIPS 2020

Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains

Meta Review

Using NTK theory the authors show that a standard multilayer perceptron fails to learn high frequencies both in theory and in practice. The authors then use a Fourier feature mapping to transform to overcome this bias. The experimental results also demonstrate that, with the same amount of training points (e,g. 1/4 of the pixels for 2D images), the reconstruction performance on 2D images and 3D shapes are significantly better using Fourier Feature Mapping than the standard MLP methods. All reviewers thought the paper contains a very rich set of experiments and interesting numerical results. The reviewers raised various technical concerns in their reviews but thought that the authors’ response adequately addressed these concerns and multiple reviewers raised their score. I concur with this assessment and recommend acceptance.