Conditional Video Generation for High-Efficiency Video Compression

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yi, Fangqiu, Xu, Jingyu, Shao, Jiawei, Zhang, Chi, Li, Xuelong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912603986460672
author Yi, Fangqiu
Xu, Jingyu
Shao, Jiawei
Zhang, Chi
Li, Xuelong
author_facet Yi, Fangqiu
Xu, Jingyu
Shao, Jiawei
Zhang, Chi
Li, Xuelong
contents Perceptual studies demonstrate that conditional diffusion models excel at reconstructing video content aligned with human visual perception. Building on this insight, we propose a video compression framework that leverages conditional diffusion models for perceptually optimized reconstruction. Specifically, we reframe video compression as a conditional generation task, where a generative model synthesizes video from sparse, yet informative signals. Our approach introduces three key modules: (1) Multi-granular conditioning that captures both static scene structure and dynamic spatio-temporal cues; (2) Compact representations designed for efficient transmission without sacrificing semantic richness; (3) Multi-condition training with modality dropout and role-aware embeddings, which prevent over-reliance on any single modality and enhance robustness. Extensive experiments show that our method significantly outperforms both traditional and neural codecs on perceptual quality metrics such as Fréchet Video Distance (FVD) and LPIPS, especially under high compression ratios.
format Preprint
id arxiv_https___arxiv_org_abs_2507_15269
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Conditional Video Generation for High-Efficiency Video Compression
Yi, Fangqiu
Xu, Jingyu
Shao, Jiawei
Zhang, Chi
Li, Xuelong
Computer Vision and Pattern Recognition
Artificial Intelligence
Perceptual studies demonstrate that conditional diffusion models excel at reconstructing video content aligned with human visual perception. Building on this insight, we propose a video compression framework that leverages conditional diffusion models for perceptually optimized reconstruction. Specifically, we reframe video compression as a conditional generation task, where a generative model synthesizes video from sparse, yet informative signals. Our approach introduces three key modules: (1) Multi-granular conditioning that captures both static scene structure and dynamic spatio-temporal cues; (2) Compact representations designed for efficient transmission without sacrificing semantic richness; (3) Multi-condition training with modality dropout and role-aware embeddings, which prevent over-reliance on any single modality and enhance robustness. Extensive experiments show that our method significantly outperforms both traditional and neural codecs on perceptual quality metrics such as Fréchet Video Distance (FVD) and LPIPS, especially under high compression ratios.
title Conditional Video Generation for High-Efficiency Video Compression
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2507.15269