SCOPE accelerates video diffusion transformers about 2x by scoring keys through 3D-RoPE subspace clusters and adaptively setting per-head Top-k counts, matching dense-attention fidelity closely.
DFSAttn: Dynamic Fine-grained Sparse Attention for Efficient Video Generation
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
Diffusion transformers have achieved remarkable success in high-quality video generation, yet their reliance on spatiotemporal 3D full attention incurs prohibitive computational cost due to the quadratic complexity of attention. Block sparse attention is a common approach to mitigate this by focusing computation on important regions. However, attention maps in DiTs exhibit inherently dynamic and fine-grained sparsity, which causes existing block sparse attention methods to degrade significantly in quality, especially at high sparsity ratios. In this paper, we revisit block sparse attention and derive a theoretical lower bound on attention recall to characterize the key factors governing its effectiveness. Guided by these insights, we propose DFSAttn, a training-free sparse attention framework that enables dynamic, fine-grained sparsification efficiently. DFSAttn incorporates three core designs: Hilbert curve-based token reordering to achieve fine-grained sparsity while preserving efficient GPU execution, hierarchical block scoring for accurate block importance estimation, and sparse mask caching with adaptive ratios to balance accuracy and efficiency. Experimental results demonstrate that DFSAttn consistently outperforms prior methods under high sparsity, achieving up to 2.1$\times$ end-to-end speedup while maintaining high generation quality. Our code is open-sourced and available at https://github.com/jessica-hujie/DFSAttn.
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SCOPE: Subspace Clustering with Online Per-Head Top-K Estimation for Sparse Video Attention
SCOPE accelerates video diffusion transformers about 2x by scoring keys through 3D-RoPE subspace clusters and adaptively setting per-head Top-k counts, matching dense-attention fidelity closely.