A Prior of a Googol Gaussians: a Tensor Ring Induced Prior for Generative Models

doi:doi:10.57702/msdhvr3u

A Prior of a Googol Gaussians: a Tensor Ring Induced Prior for Generative Models

Generative models produce realistic objects in many domains, including text, image, video, and audio synthesis. Most popular models—Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs)—usually employ a standard Gaussian distribution as a prior. Previous works show that the richer family of prior distributions may help to avoid the mode collapse problem in GANs and to improve the evidence lower bound in VAEs.

BibTex:

@dataset{Maksim_Kuznetsov_and_Daniil_Polykovskiy_and_Dmitry_Vetrov_and_Alexander_Zhebrak_2024,
    abstract = {Generative models produce realistic objects in many domains, including text, image, video, and audio synthesis. Most popular models—Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs)—usually employ a standard Gaussian distribution as a prior. Previous works show that the richer family of prior distributions may help to avoid the mode collapse problem in GANs and to improve the evidence lower bound in VAEs.},
    author = {Maksim Kuznetsov and Daniil Polykovskiy and Dmitry Vetrov and Alexander Zhebrak},
    doi = {10.57702/msdhvr3u},
    institution = {No Organization},
    keyword = {'Generative Models', 'Prior Distributions', 'Tensor Ring'},
    month = {dec},
    publisher = {TIB},
    title = {A Prior of a Googol Gaussians: a Tensor Ring Induced Prior for Generative Models},
    url = {https://service.tib.eu/ldmservice/dataset/a-prior-of-a-googol-gaussians--a-tensor-ring-induced-prior-for-generative-models},
    year = {2024}
}