DiSCo compresses video by transmitting semantic text, spatiotemporally degraded frames, and motion cues, then reconstructs high-quality output via conditional video diffusion, outperforming baselines by 2-10X on perceptual metrics at low bitrates.
To- wards accurate generative models of video: A new metric & challenges
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2representative citing papers
citing papers explorer
-
Low-Bitrate Video Compression through Semantic-Conditioned Diffusion
DiSCo compresses video by transmitting semantic text, spatiotemporally degraded frames, and motion cues, then reconstructs high-quality output via conditional video diffusion, outperforming baselines by 2-10X on perceptual metrics at low bitrates.
- FlowC2S: Flowing from Current to Succeeding Frames for Fast and Memory-Efficient Video Continuation