Enhancing VVC with deep learning based multi-frame post-processing

Duolikun Danier; Chen Feng; Fan Zhang; David Bull

doi:10.48550/arXiv.2205.09458

Enhancing VVC with deep learning based multi-frame post-processing

Duolikun Danier, Chen Feng, Fan Zhang, David Bull

Research output: Contribution to conference › Paper › peer-review

Abstract / Description of output

This paper describes a CNN-based multi-frame post-processing approach based on a perceptually-inspired Generative Adversarial Network architecture, CVEGAN. This method has been integrated with the Versatile Video Coding Test Model (VTM) 15.2 to enhance the visual quality of the final reconstructed content. The evaluation results on the CLIC 2022 validation sequences show consistent coding gains over the original VVC VTM at the same bitrates when assessed by PSNR. The integrated codec has been submitted to the Challenge on Learned Image Compression (CLIC) 2022 (video track), and the team name associated with this submission is BVI_VC.

Original language	English
Pages	1-4
Number of pages	4
DOIs	https://doi.org/10.48550/arXiv.2205.09458
Publication status	Published - 19 Jun 2022
Event	5th Workshop and Challenge on Learned Image Compression - Online Duration: 19 Jun 2022 → 19 Jun 2022

Workshop

Workshop	5th Workshop and Challenge on Learned Image Compression
Period	19/06/22 → 19/06/22

Access to Document

10.48550/arXiv.2205.09458Licence: Creative Commons: Attribution-NonCommercial-ShareAlike (CC BY-NC-SA)

DanierEtalCLIC2022EnhancingVVCwithDeepLearningbasedAccepted author manuscript, 3.38 MBLicence: Creative Commons: Attribution-NonCommercial-ShareAlike (CC BY-NC-SA)

https://clic.compression.cc/2022/publications/9/index.html

Cite this

@conference{df26db5db59a47aaaf47e6b327af4dba,

title = "Enhancing VVC with deep learning based multi-frame post-processing",

abstract = "This paper describes a CNN-based multi-frame post-processing approach based on a perceptually-inspired Generative Adversarial Network architecture, CVEGAN. This method has been integrated with the Versatile Video Coding Test Model (VTM) 15.2 to enhance the visual quality of the final reconstructed content. The evaluation results on the CLIC 2022 validation sequences show consistent coding gains over the original VVC VTM at the same bitrates when assessed by PSNR. The integrated codec has been submitted to the Challenge on Learned Image Compression (CLIC) 2022 (video track), and the team name associated with this submission is BVI_VC.",

author = "Duolikun Danier and Chen Feng and Fan Zhang and David Bull",

year = "2022",

month = jun,

day = "19",

doi = "10.48550/arXiv.2205.09458",

language = "English",

pages = "1--4",

note = "5th Workshop and Challenge on Learned Image Compression ; Conference date: 19-06-2022 Through 19-06-2022",

}

TY - CONF

T1 - Enhancing VVC with deep learning based multi-frame post-processing

AU - Danier, Duolikun

AU - Feng, Chen

AU - Zhang, Fan

AU - Bull, David

PY - 2022/6/19

Y1 - 2022/6/19

N2 - This paper describes a CNN-based multi-frame post-processing approach based on a perceptually-inspired Generative Adversarial Network architecture, CVEGAN. This method has been integrated with the Versatile Video Coding Test Model (VTM) 15.2 to enhance the visual quality of the final reconstructed content. The evaluation results on the CLIC 2022 validation sequences show consistent coding gains over the original VVC VTM at the same bitrates when assessed by PSNR. The integrated codec has been submitted to the Challenge on Learned Image Compression (CLIC) 2022 (video track), and the team name associated with this submission is BVI_VC.

AB - This paper describes a CNN-based multi-frame post-processing approach based on a perceptually-inspired Generative Adversarial Network architecture, CVEGAN. This method has been integrated with the Versatile Video Coding Test Model (VTM) 15.2 to enhance the visual quality of the final reconstructed content. The evaluation results on the CLIC 2022 validation sequences show consistent coding gains over the original VVC VTM at the same bitrates when assessed by PSNR. The integrated codec has been submitted to the Challenge on Learned Image Compression (CLIC) 2022 (video track), and the team name associated with this submission is BVI_VC.

U2 - 10.48550/arXiv.2205.09458

DO - 10.48550/arXiv.2205.09458

M3 - Paper

SP - 1

EP - 4

T2 - 5th Workshop and Challenge on Learned Image Compression

Y2 - 19 June 2022 through 19 June 2022

ER -

Enhancing VVC with deep learning based multi-frame post-processing

Abstract / Description of output

Workshop

Access to Document

Fingerprint

Cite this