Backpropagating through Fréchet Inception Distance
arXiv:2009.14075v26 citations
AI Analysis
This work addresses the challenge of optimizing generative models for better evaluation metrics, but it is incremental as it builds on existing FID usage.
The paper tackled the problem of training generative models more effectively by using the Fréchet Inception Distance (FID) as a loss function, resulting in improved FID scores for Generative Adversarial Networks.
The Fréchet Inception Distance (FID) has been used to evaluate hundreds of generative models. We introduce FastFID, which can efficiently train generative models with FID as a loss function. Using FID as an additional loss for Generative Adversarial Networks improves their FID.