We propose a deep residual autoencoder exploiting Residual-in-Residual Dense Blocks (RRDB) leveraging both the learning capacity of deep residual networks and prior knowledge of the JPEG compression pipeline. The proposed method is blind and universal, consisting of a unique model that effectively restores images with any level of compression. It operates in the YCbCr color space and performs JPEG restoration in two phases using two different autoencoders: the first one restores the luma channel exploiting 2D convolutions; the second one, using the restored luma channel as a guide, restores the chroma channels exploiting 3D convolutions. Extensive experimental results on four widely used benchmark datasets (i.e. LIVE1, BDS500, CLASSIC-5, and Kodak) show that our model outperforms state of the art methods, even those using a different set of weights for each compression quality, in terms of all the evaluation metrics considered (i.e. PSNR, PSNR-B, and SSIM). Furthermore, the proposed model shows a greater robustness than state-of-the-art methods when applied to compression qualities not seen during training.

Zini, S., Bianco, S., Schettini, R. (2020). Deep Residual Autoencoder for Blind Universal JPEG Restoration. IEEE ACCESS, 8, 63283-63294 [10.1109/ACCESS.2020.2984387].

Deep Residual Autoencoder for Blind Universal JPEG Restoration

Zini S.;Bianco S.;Schettini R.
2020

Abstract

We propose a deep residual autoencoder exploiting Residual-in-Residual Dense Blocks (RRDB) leveraging both the learning capacity of deep residual networks and prior knowledge of the JPEG compression pipeline. The proposed method is blind and universal, consisting of a unique model that effectively restores images with any level of compression. It operates in the YCbCr color space and performs JPEG restoration in two phases using two different autoencoders: the first one restores the luma channel exploiting 2D convolutions; the second one, using the restored luma channel as a guide, restores the chroma channels exploiting 3D convolutions. Extensive experimental results on four widely used benchmark datasets (i.e. LIVE1, BDS500, CLASSIC-5, and Kodak) show that our model outperforms state of the art methods, even those using a different set of weights for each compression quality, in terms of all the evaluation metrics considered (i.e. PSNR, PSNR-B, and SSIM). Furthermore, the proposed model shows a greater robustness than state-of-the-art methods when applied to compression qualities not seen during training.
Articolo in rivista - Articolo scientifico
autoencoder; deep learning; JPEG restoration; residual network
English
2020
8
63283
63294
9050792
open
Zini, S., Bianco, S., Schettini, R. (2020). Deep Residual Autoencoder for Blind Universal JPEG Restoration. IEEE ACCESS, 8, 63283-63294 [10.1109/ACCESS.2020.2984387].
File in questo prodotto:
File Dimensione Formato  
10281-278942_VoR.pdf

accesso aperto

Tipologia di allegato: Publisher’s Version (Version of Record, VoR)
Licenza: Creative Commons
Dimensione 2.02 MB
Formato Adobe PDF
2.02 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/278942
Citazioni
  • Scopus 16
  • ???jsp.display-item.citation.isi??? 11
Social impact