Using Skew to Assess the Quality of GAN-generated Image Features
Abstract: The rapid advancement of Generative Adversarial Networks (GANs) necessitates the need to robustly evaluate these models. Among the established evaluation criteria, the Fréchet Inception Distance (FID) has been widely adopted due to its conceptual simplicity, fast computation time, and strong correlation with human perception. However, FID has inherent limitations, mainly stemming from its assumption that feature embeddings follow a Gaussian distribution, and therefore can be defined by their first two moments. As this does not hold in practice, in this paper we explore the importance of third-moments in image feature data and use this information to define a new measure, which we call the Skew Inception Distance (SID). We prove that SID is a pseudometric on probability distributions, show how it extends FID, and present a practical method for its computation. Our numerical experiments support that SID either tracks with FID or, in some cases, aligns more closely with human perception when evaluating image features of ImageNet data. Our work also shows that principal component analysis can be used to speed up the computation time of both FID and SID. Although we focus on using SID on image features for GAN evaluation, SID is applicable much more generally, including for the evaluation of other generative models.
Submission Length: Regular submission (no more than 12 pages of main content)
Changes Since Last Submission: In the revision, we have added experiments that study SID using different models for feature extraction and different datasets. We also clarified the computational complexity of SID, and incorporated reviewer suggestions about limitations and broader impact into the Conclusion.
Assigned Action Editor: ~Masha_Itkina1
Submission Number: 1900
Loading