Explicitly Minimizing the Blur Error of Variational Autoencoders


Loading...

Date

2023

Publication Type

Conference Paper

ETH Bibliography

yes

Citations

Altmetric

Data

Abstract

Variational autoencoders (VAEs) are powerful generative modelling methods, however they suffer from blurry generated samples and reconstructions compared to the images they have been trained on. Significant research effort has been spent to increase the generative capabilities by creating more flexible models but often flexibility comes at the cost of higher complexity and computational cost. Several works have focused on altering the reconstruction term of the evidence lower bound (ELBO), however, often at the expense of losing the mathematical link to maximizing the likelihood of the samples under the modeled distribution. Here we propose a new formulation of the reconstruction term for the VAE that specifically penalizes the generation of blurry images while at the same time still maximizing the ELBO under the modeled distribution. We show the potential of the proposed loss on three different data sets, where it outperforms several recently proposed reconstruction losses for VAEs.

Publication status

published

Editor

Book title

The Eleventh International Conference on Learning Representations

Journal / series

Volume

Pages / Article No.

Publisher

OpenReview

Event

11th International Conference on Learning Representations (ICLR 2023)

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Variational Autoencoders; Generative Modelling; Blur

Organisational unit

09579 - Konukoglu, Ender / Konukoglu, Ender check_circle

Notes

Funding

Related publications and datasets