Evaluating generative networks using Gaussian mixtures of image features
This work addresses the evaluation of generative networks for researchers, offering a more robust metric, but it is incremental as it builds on FID by improving its statistical assumptions.
The authors tackled the problem that the Fréchet Inception Distance (FID) assumes Gaussian distributions for image features, which they show does not hold for ImageNet, by proposing WaM, a new metric using Gaussian mixture models and restricted 2-Wasserstein distance, and experimentally demonstrated that WaM is less sensitive to imperceptible perturbations than FID.
We develop a measure for evaluating the performance of generative networks given two sets of images. A popular performance measure currently used to do this is the Fréchet Inception Distance (FID). FID assumes that images featurized using the penultimate layer of Inception-v3 follow a Gaussian distribution, an assumption which cannot be violated if we wish to use FID as a metric. However, we show that Inception-v3 features of the ImageNet dataset are not Gaussian; in particular, every single marginal is not Gaussian. To remedy this problem, we model the featurized images using Gaussian mixture models (GMMs) and compute the 2-Wasserstein distance restricted to GMMs. We define a performance measure, which we call WaM, on two sets of images by using Inception-v3 (or another classifier) to featurize the images, estimate two GMMs, and use the restricted $2$-Wasserstein distance to compare the GMMs. We experimentally show the advantages of WaM over FID, including how FID is more sensitive than WaM to imperceptible image perturbations. By modelling the non-Gaussian features obtained from Inception-v3 as GMMs and using a GMM metric, we can more accurately evaluate generative network performance.