OpenDCVCs: A PyTorch Open Source Implementation and Performance Evaluation of the DCVC series Video Codecs

TL;DR

OpenDCVCs is an open-source PyTorch framework that enables reproducible research, training, and benchmarking of advanced learned video compression models, significantly aiding community development and comparison.

Contribution

It provides the first comprehensive, training-ready implementation of multiple DCVC models, facilitating reproducibility, benchmarking, and further research in learned video compression.

Findings

DCVC models outperform classical codecs in bitrate reduction

OpenDCVCs enables consistent evaluation across datasets

Framework supports end-to-end training and benchmarking

Abstract

We present OpenDCVCs, an open-source PyTorch implementation designed to advance reproducible research in learned video compression. OpenDCVCs provides unified and training-ready implementations of four representative Deep Contextual Video Compression (DCVC) models--DCVC, DCVC with Temporal Context Modeling (DCVC-TCM), DCVC with Hybrid Entropy Modeling (DCVC-HEM), and DCVC with Diverse Contexts (DCVC-DC). While the DCVC series achieves substantial bitrate reductions over both classical codecs and advanced learned models, previous public code releases have been limited to evaluation codes, presenting significant barriers to reproducibility, benchmarking, and further development. OpenDCVCs bridges this gap by offering a comprehensive, self-contained framework that supports both end-to-end training and evaluation for all included algorithms. The implementation includes detailed documentation, evaluation protocols, and extensive benchmarking results across diverse datasets, providing a transparent and consistent foundation for comparison and extension. All code and experimental tools are publicly available at https://gitlab.com/viper-purdue/opendcvcs, empowering the community to accelerate research and foster collaboration.