Can No-Reference Quality-Assessment Methods Serve as Perceptual Losses for Super-Resolution?

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kashkarov, Egor, Chistov, Egor, Molodetskikh, Ivan, Vatolin, Dmitriy
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917679688843264
author Kashkarov, Egor
Chistov, Egor
Molodetskikh, Ivan
Vatolin, Dmitriy
author_facet Kashkarov, Egor
Chistov, Egor
Molodetskikh, Ivan
Vatolin, Dmitriy
contents Perceptual losses play an important role in constructing deep-neural-network-based methods by increasing the naturalness and realism of processed images and videos. Use of perceptual losses is often limited to LPIPS, a fullreference method. Even though deep no-reference image-qualityassessment methods are excellent at predicting human judgment, little research has examined their incorporation in loss functions. This paper investigates direct optimization of several video-superresolution models using no-reference image-quality-assessment methods as perceptual losses. Our experimental results show that straightforward optimization of these methods produce artifacts, but a special training procedure can mitigate them.
format Preprint
id arxiv_https___arxiv_org_abs_2405_20392
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Can No-Reference Quality-Assessment Methods Serve as Perceptual Losses for Super-Resolution?
Kashkarov, Egor
Chistov, Egor
Molodetskikh, Ivan
Vatolin, Dmitriy
Image and Video Processing
Computer Vision and Pattern Recognition
Perceptual losses play an important role in constructing deep-neural-network-based methods by increasing the naturalness and realism of processed images and videos. Use of perceptual losses is often limited to LPIPS, a fullreference method. Even though deep no-reference image-qualityassessment methods are excellent at predicting human judgment, little research has examined their incorporation in loss functions. This paper investigates direct optimization of several video-superresolution models using no-reference image-quality-assessment methods as perceptual losses. Our experimental results show that straightforward optimization of these methods produce artifacts, but a special training procedure can mitigate them.
title Can No-Reference Quality-Assessment Methods Serve as Perceptual Losses for Super-Resolution?
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2405.20392