Parametric Adversarial Divergences are Good Task Losses for Generative Modeling

13 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.

Adversarial Example Games (2020, neutral): “This alternative perspective on the game (AEG) shares similarities with the divergence minimization perspective on GANs [40].”
A Variational Inequality Perspective on Generative Adversarial Networks (2018, neutral): “Most of them can be formulated as “parametric adversarial divergences” (Huang et al., 2017): Div(p||qθ) def = sup φ∈Φ E(x,x′)∼p⊗qθ [∆(fφ(x), fφ(x ′))] , (1)”
Boundary-Seeking Generative Adversarial Networks (2017, background): “Maximizing Equation 5 provides a neural estimator of f -divergence, or neural divergence (Huang et al., 2018).”
Minimax Theorem for Latent Games or: How I Learned to Stop Worrying about Mixed-Nash and Love Neural Nets (2020, background): “Similarly, in the context of Generative Adversarial Networks, it has been argued that setting a restricted function class for the discriminator provides a more meaningful loss and describes an achievable learning task for the generator Arora et al. (2017); Huang et al. (2017).”
A Variational Inequality Perspective on Generative Adversarial Nets (2018, background): “Most of the parametric divergences encountered in the GAN literature can be formulated as so-called parametric adversarial divergences (Huang et al., 2017): ar X iv :1 80 2.”