Ultra Flash: Scaling Real-Time Streaming Video Generation to High Resolutions

arXiv:2606.09150 · cs.CV · Submitted 2026-06-08 · Read on arXiv

cs.CV

Submitted: 2026-06-08

Updated: 2026-10-13

Code: https://github.com/madebyollin/taehv

Project page: https://xin1u.github.io/UltraFlash/1

Terminology

• Computational overhead — 1× in this paper · explained in 2 episodes
• Diffusion Model — 1× in this paper · explained in 9 episodes
• Diffusion Models — 1× in this paper · explained in 6 episodes
• Distillation — 1× in this paper · explained in 2 episodes
• Generative Model — 1× in this paper · explained in 2 episodes
• Generative Models — 1× in this paper · explained in 3 episodes

Sources

• Longformer: The Long-Document Transformer
• Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets
• ResAdapter: Domain Consistent Resolution Adapter for Diffusion Models
• Generating Long Sequences with Sparse Transformers
• Self-Forcing++: Towards Minute-Scale High-Quality Video Generation
• DemoFusion: Democratising High-Resolution Image Generation With No $$$
• FSVideo: Fast Speed Video Diffusion Model in a Highly-Compressed Latent Space
• Seedance 1.0: Exploring the Boundaries of Video Generation Models
• Efficient Autoregressive Video Diffusion with Dummy Head
• AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models without Specific Tuning
• LTX-Video: Realtime Video Latent Diffusion
• LTX-2: Efficient Joint Audio-Visual Foundation Model
• VEnhancer: Generative Space-Time Enhancement for Video Generation
• ScaleCrafter: Tuning-free Higher-Resolution Visual Generation with Diffusion Models
• Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion
• Pyramidal Flow Matching for Efficient Video Generative Modeling
• FIFO-Diffusion: Generating Infinite Videos from Text without Training
• HunyuanVideo: A Systematic Framework For Large Video Generative Models
• Decoupled DMD: CFG Augmentation as the Spear, Distribution Matching as the Shield
• Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models