A Domain Agnostic Measure for Monitoring and Evaluating GANs

By Paulina Grnarova, Kfir Yehuda Levy, Aurélien Lucchi, Nathanael Perraudin, Ian J. Goodfellow, Thomas Hofmann, Andreas Krause. 2018

41 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Ngoc-Trung Tran
2. Sam Bond-Taylor
3. Ngai-Man Cheung
4. Kwot Sin Lee
5. Ali Borji
6. Adam Leach
7. Yang Long
8. Chris G. Willcocks

Citations in the corpus

• InfoMax-GAN: Improved Adversarial Image Generation via Information Maximization and Contrastive Learning (2020, background): “These works instead focus on direct metrics [7], [8], [41], [54]–[56] for assessing the diversity and quality of generated images.”
• Pros and Cons of GAN Evaluation Measures: New Developments (2021, background): “Grnarova et al. (2019) leverage the notion of duality gap from game theory to propose a domain-agnostic and computationally-efficient measure that can be used to assess different models as well as monitoring the progress of a single model throughout training.”
• Mimicry: Towards the Reproducibility of GAN Research (2020, background): “As future work, we aim to cover a wider range of GANs and metrics [22, 34, 13], and expand to include other GAN tasks, such as 3D view synthesis.”
• Deep Generative Modelling: A Comparative Review of VAEs, GANs, Normalizing Flows, Energy-Based and Autoregressive Models (2021, background): “Originally proposed as a proxy to measure GAN convergence [68], the duality gap is an upper bound on the JS-”