Versatile Recompression-Aware Perceptual Image Super-Resolution

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: He, Mingwei, Xu, Tongda, Ge, Xingtong, Sun, Ming, Zhou, Chao, Wang, Yan
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910077354508288
author He, Mingwei
Xu, Tongda
Ge, Xingtong
Sun, Ming
Zhou, Chao
Wang, Yan
author_facet He, Mingwei
Xu, Tongda
Ge, Xingtong
Sun, Ming
Zhou, Chao
Wang, Yan
contents Perceptual image super-resolution (SR) methods restore degraded images and produce sharp outputs. In practice, those outputs are usually recompressed for storage and transmission. Ignoring recompression is suboptimal as the downstream codec might add additional artifacts to restored images. However, jointly optimizing SR and recompression is challenging, as the codecs are not differentiable and vary in configuration. In this paper, we present \textbf{Versatile Recompression-Aware Perceptual Super-Resolution (VRPSR)}, which makes existing perceptual SR aware of versatile compression. First, we formulate compression as conditional text-to-image generation and utilize a pre-trained diffusion model to build a generalizable codec simulator. Next, we propose a set of training techniques tailored for perceptual SR, including optimizing the simulator using perceptual targets and adopting slightly compressed images as the training target. Empirically, our VRPSR achieves 10% - 40% bitrate savings based on Real-ESRGAN and S3Diff under H.264/H.265/H.266 single-picture (intra) compression. Besides, our VRPSR facilitates joint optimization of SR and the post-processing model after recompression.
format Preprint
id arxiv_https___arxiv_org_abs_2511_18090
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Versatile Recompression-Aware Perceptual Image Super-Resolution
He, Mingwei
Xu, Tongda
Ge, Xingtong
Sun, Ming
Zhou, Chao
Wang, Yan
Computer Vision and Pattern Recognition
Perceptual image super-resolution (SR) methods restore degraded images and produce sharp outputs. In practice, those outputs are usually recompressed for storage and transmission. Ignoring recompression is suboptimal as the downstream codec might add additional artifacts to restored images. However, jointly optimizing SR and recompression is challenging, as the codecs are not differentiable and vary in configuration. In this paper, we present \textbf{Versatile Recompression-Aware Perceptual Super-Resolution (VRPSR)}, which makes existing perceptual SR aware of versatile compression. First, we formulate compression as conditional text-to-image generation and utilize a pre-trained diffusion model to build a generalizable codec simulator. Next, we propose a set of training techniques tailored for perceptual SR, including optimizing the simulator using perceptual targets and adopting slightly compressed images as the training target. Empirically, our VRPSR achieves 10% - 40% bitrate savings based on Real-ESRGAN and S3Diff under H.264/H.265/H.266 single-picture (intra) compression. Besides, our VRPSR facilitates joint optimization of SR and the post-processing model after recompression.
title Versatile Recompression-Aware Perceptual Image Super-Resolution
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.18090