DecoupAttn: Decoupled Sparse Attention for Training-Free Acceleration of Video Diffusion
cs.CV
Submitted: 2026-11-04
Updated: 2026-11-04
Terminology
Sources
- MC-Sparse: Deconstructing and Closing the Dense-Sparse Attention Gap in Diffusion Transformers
- Generative World Renderer
- HunyuanVideo: A Systematic Framework For Large Video Generative Models
- HyperVAttention: Efficient Sparse Attention with Spatio-Temporal Clustering for Video Diffusion
- Sol-Attn: Accelerating Video Generation Inference via On-the-Fly Attention Sparsification
- PISA: Piecewise Sparse Attention Is Wiser for Efficient Diffusion Transformers
- MomentKV: Closing the Directional Gap in KV Cache Eviction for Long-Context Inference
- LoSA: Near-Lossless Sparse Attention for Training-Free Video Diffusion Acceleration
- Rectified SpaAttn: Revisiting Attention Sparsity for Efficient Video Generation
- Mixture of Distributions Matters: Dynamic Sparse Attention for Efficient Video Diffusion Transformers
- Cosmos 3: Omnimodal World Models for Physical AI
- Cosmos World Foundation Model Platform for Physical AI
- World Simulation with Video Foundation Models for Physical AI
- LiteAttention: A Temporal Sparse Attention for Diffusion Transformers
- Partition the Support, Reconstruct the Residual: Training-Free Sparse Attention for Video Generation and World Models
- Wan: Open and Advanced Large-Scale Video Generative Models
- Sparsity Should Be ALLOTted by Output: Training-Free Allocation of Block-Sparse Attention for Video Diffusion Transformers
- RECAP: Reciprocal Sparse Attention and Hierarchical Caching for Efficient Video Diffusion
- Accelerating Text-to-Video Generation with Calibrated Sparse Attention
- Ride the Wave: Precision-Allocated Sparse Attention for Smooth Video Generation
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs