Uncertainty Estimation for Super-Resolution using ESRGAN

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Adapa, Maniraj Sai, Zullich, Marco, Valdenegro-Toro, Matias
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909435813691392
author Adapa, Maniraj Sai
Zullich, Marco
Valdenegro-Toro, Matias
author_facet Adapa, Maniraj Sai
Zullich, Marco
Valdenegro-Toro, Matias
contents Deep Learning-based image super-resolution (SR) has been gaining traction with the aid of Generative Adversarial Networks. Models like SRGAN and ESRGAN are constantly ranked between the best image SR tools. However, they lack principled ways for estimating predictive uncertainty. In the present work, we enhance these models using Monte Carlo-Dropout and Deep Ensemble, allowing the computation of predictive uncertainty. When coupled with a prediction, uncertainty estimates can provide more information to the model users, highlighting pixels where the SR output might be uncertain, hence potentially inaccurate, if these estimates were to be reliable. Our findings suggest that these uncertainty estimates are decently calibrated and can hence fulfill this goal, while providing no performance drop with respect to the corresponding models without uncertainty estimation.
format Preprint
id arxiv_https___arxiv_org_abs_2412_15439
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Uncertainty Estimation for Super-Resolution using ESRGAN
Adapa, Maniraj Sai
Zullich, Marco
Valdenegro-Toro, Matias
Image and Video Processing
Computer Vision and Pattern Recognition
Deep Learning-based image super-resolution (SR) has been gaining traction with the aid of Generative Adversarial Networks. Models like SRGAN and ESRGAN are constantly ranked between the best image SR tools. However, they lack principled ways for estimating predictive uncertainty. In the present work, we enhance these models using Monte Carlo-Dropout and Deep Ensemble, allowing the computation of predictive uncertainty. When coupled with a prediction, uncertainty estimates can provide more information to the model users, highlighting pixels where the SR output might be uncertain, hence potentially inaccurate, if these estimates were to be reliable. Our findings suggest that these uncertainty estimates are decently calibrated and can hence fulfill this goal, while providing no performance drop with respect to the corresponding models without uncertainty estimation.
title Uncertainty Estimation for Super-Resolution using ESRGAN
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2412.15439