Ultra-low bitrate video conferencing using deep image animation - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2021

Ultra-low bitrate video conferencing using deep image animation

Résumé

In this work we propose a novel deep learning approach for ultra-low bitrate video compression for video conferencing applications. To address the shortcomings of current video compression paradigms when the available bandwidth is extremely limited, we adopt a model-based approach that employs deep neural networks to encode motion information as keypoint displacement and reconstruct the video signal at the decoder side. The overall system is trained in an end-to-end fashion minimizing a reconstruction error on the encoder output. Objective and subjective quality evaluation experiments demonstrate that the proposed approach provides an average bitrate reduction for the same visual quality of more than 80% compared to HEVC.
Fichier principal
Vignette du fichier
Icassp_deep_Comp.pdf (963.54 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03138045 , version 1 (26-02-2021)
hal-03138045 , version 2 (29-12-2021)

Identifiants

Citer

Goluck Konuko, Giuseppe Valenzise, Stéphane Lathuilière. Ultra-low bitrate video conferencing using deep image animation. ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Jun 2021, Toronto, Canada. ⟨10.1109/icassp39728.2021.9414731⟩. ⟨hal-03138045v2⟩
100 Consultations
272 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More