Low-Bitrate Video Compression through Semantic-Conditioned Diffusion

TL;DR

The paper introduces DiSCo, a semantic video compression method that uses minimal data and generative models to produce high-quality videos at ultra-low bitrates, outperforming traditional codecs.

Contribution

DiSCo is a novel semantic compression framework that decomposes videos into semantic, appearance, and motion cues, enabling efficient transmission and high-quality reconstruction.

Findings

Outperforms baseline codecs by 2-10X on perceptual metrics at low bitrates.

Uses a conditional diffusion model for high-quality, temporally coherent video synthesis.

Employs multimodal representations including text, degraded video, and sketches or poses.

Abstract

Traditional video codecs optimized for pixel fidelity collapse at ultra-low bitrates and produce severe artifacts. This failure arises from a fundamental misalignment between pixel accuracy and human perception. We propose a semantic video compression framework named DiSCo that transmits only the most meaningful information while relying on generative priors for detail synthesis. The source video is decomposed into three compact modalities: a textual description, a spatiotemporally degraded video, and optional sketches or poses that respectively capture semantic, appearance, and motion cues. A conditional video diffusion model then reconstructs high-quality, temporally coherent videos from these compact representations. Temporal forward filling, token interleaving, and modality-specific codecs are proposed to improve multimodal generation and modality compactness. Experiments show that our method outperforms baseline semantic and traditional codecs by 2-10X on perceptual metrics at low bitrates.