Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | 2021 IEEE International Conference on Image Processing, ICIP 2021 |
| Herausgeber (Verlag) | IEEE Computer Society |
| Seiten | 2114-2118 |
| Seitenumfang | 5 |
| ISBN (elektronisch) | 9781665441155 |
| ISBN (Print) | 978-1-6654-3102-6 |
| Publikationsstatus | Veröffentlicht - 2021 |
| Veranstaltung | 2021 IEEE International Conference on Image Processing, ICIP 2021 - Anchorage, USA / Vereinigte Staaten Dauer: 19 Sept. 2021 → 22 Sept. 2021 |
Publikationsreihe
| Name | Proceedings - International Conference on Image Processing, ICIP |
|---|---|
| Band | 2021-September |
| ISSN (Print) | 1522-4880 |
Abstract
In this paper we introduce an error concealment method for VVC based on deep recurrent neural networks, which employs the PredNet model to estimate missing video frames by using past decoded frames. The network is trained using the BVI-DVC data set to infer even full-HD frames. We integrated our proposed model in the VVC reference software VTM for its evaluation. It performs, in average, 6 dB or up to 5 dB better than the frame copy model in terms of PSNR measurements for a concealed I-frame or P-frame, respectively.
ASJC Scopus Sachgebiete
- Informatik (insg.)
- Software
- Informatik (insg.)
- Maschinelles Sehen und Mustererkennung
- Informatik (insg.)
- Signalverarbeitung
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
2021 IEEE International Conference on Image Processing, ICIP 2021. IEEE Computer Society, 2021. S. 2114-2118 (Proceedings - International Conference on Image Processing, ICIP; Band 2021-September).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - NEURAL NETWORK-BASED ERROR CONCEALMENT FOR VVC
AU - Benjak, Martin
AU - Samayoa, Yasser
AU - Ostermann, Jörn
PY - 2021
Y1 - 2021
N2 - In this paper we introduce an error concealment method for VVC based on deep recurrent neural networks, which employs the PredNet model to estimate missing video frames by using past decoded frames. The network is trained using the BVI-DVC data set to infer even full-HD frames. We integrated our proposed model in the VVC reference software VTM for its evaluation. It performs, in average, 6 dB or up to 5 dB better than the frame copy model in terms of PSNR measurements for a concealed I-frame or P-frame, respectively.
AB - In this paper we introduce an error concealment method for VVC based on deep recurrent neural networks, which employs the PredNet model to estimate missing video frames by using past decoded frames. The network is trained using the BVI-DVC data set to infer even full-HD frames. We integrated our proposed model in the VVC reference software VTM for its evaluation. It performs, in average, 6 dB or up to 5 dB better than the frame copy model in terms of PSNR measurements for a concealed I-frame or P-frame, respectively.
KW - Error concealment
KW - Video coding
KW - Video communication
KW - VVC
UR - http://www.scopus.com/inward/record.url?scp=85125596598&partnerID=8YFLogxK
U2 - 10.1109/ICIP42928.2021.9506399
DO - 10.1109/ICIP42928.2021.9506399
M3 - Conference contribution
AN - SCOPUS:85125596598
SN - 978-1-6654-3102-6
T3 - Proceedings - International Conference on Image Processing, ICIP
SP - 2114
EP - 2118
BT - 2021 IEEE International Conference on Image Processing, ICIP 2021
PB - IEEE Computer Society
T2 - 2021 IEEE International Conference on Image Processing, ICIP 2021
Y2 - 19 September 2021 through 22 September 2021
ER -