Enregistré dans:
Détails bibliographiques
Auteurs principaux: Gao, Ge, Kwan, Ho Man, Zhang, Fan, Bull, David
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:https://arxiv.org/abs/2409.00953
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866909302482010112
author Gao, Ge
Kwan, Ho Man
Zhang, Fan
Bull, David
author_facet Gao, Ge
Kwan, Ho Man
Zhang, Fan
Bull, David
contents Neural video compression has recently demonstrated significant potential to compete with conventional video codecs in terms of rate-quality performance. These learned video codecs are however associated with various issues related to decoding complexity (for autoencoder-based methods) and/or system delays (for implicit neural representation (INR) based models), which currently prevent them from being deployed in practical applications. In this paper, targeting a practical neural video codec, we propose a novel INR-based coding framework, PNVC, which innovatively combines autoencoder-based and overfitted solutions. Our approach benefits from several design innovations, including a new structural reparameterization-based architecture, hierarchical quality control, modulation-based entropy modeling, and scale-aware positional embedding. Supporting both low delay (LD) and random access (RA) configurations, PNVC outperforms existing INR-based codecs, achieving nearly 35%+ BD-rate savings against HEVC HM 18.0 (LD) - almost 10% more compared to one of the state-of-the-art INR-based codecs, HiNeRV and 5% more over VTM 20.0 (LD), while maintaining 20+ FPS decoding speeds for 1080p content. This represents an important step forward for INR-based video coding, moving it towards practical deployment. The source code will be available for public evaluation.
format Preprint
id arxiv_https___arxiv_org_abs_2409_00953
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PNVC: Towards Practical INR-based Video Compression
Gao, Ge
Kwan, Ho Man
Zhang, Fan
Bull, David
Computer Vision and Pattern Recognition
Neural video compression has recently demonstrated significant potential to compete with conventional video codecs in terms of rate-quality performance. These learned video codecs are however associated with various issues related to decoding complexity (for autoencoder-based methods) and/or system delays (for implicit neural representation (INR) based models), which currently prevent them from being deployed in practical applications. In this paper, targeting a practical neural video codec, we propose a novel INR-based coding framework, PNVC, which innovatively combines autoencoder-based and overfitted solutions. Our approach benefits from several design innovations, including a new structural reparameterization-based architecture, hierarchical quality control, modulation-based entropy modeling, and scale-aware positional embedding. Supporting both low delay (LD) and random access (RA) configurations, PNVC outperforms existing INR-based codecs, achieving nearly 35%+ BD-rate savings against HEVC HM 18.0 (LD) - almost 10% more compared to one of the state-of-the-art INR-based codecs, HiNeRV and 5% more over VTM 20.0 (LD), while maintaining 20+ FPS decoding speeds for 1080p content. This represents an important step forward for INR-based video coding, moving it towards practical deployment. The source code will be available for public evaluation.
title PNVC: Towards Practical INR-based Video Compression
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2409.00953