paper-with-me

홈 › Papers

HASTE: Training-Free Video Diffusion Acceleration via Head-Wise Adaptive Sparse Attention

2026-05-14 · Xuzhe Zheng, Yuexiao Ma, Jing Xu, Xiawu Zheng, Rongrong Ji, Fei Chao arxiv

Diffusion-based video generation has advanced substantially in visual fidelity and temporal coherence, but practical deployment remains limited by the quadratic complexity of full attention. Training-free sparse attention is attractive because it accelerates pretrained models without retraining, yet existing online top-$p$ sparse attention still spends non-negligible cost on mask prediction and applies shared thresholds despite strong head-level heterogeneity. We show that these two overlooked factors limit the practical speed-quality trade-off of training-free sparse attention in Video DiTs. To address them, we introduce a head-wise adaptive framework with two plug-in components: Temporal Mask Reuse, which skips unnecessary mask prediction based on query-key drift, and Error-guided Budgeted Calibration, which assigns per-head top-$p$ thresholds by minimizing measured model-output error under a global sparsity budget. On Wan2.1-1.3B and Wan2.1-14B, our method consistently improves XAttention and SVG2, achieving up to 1.93 times speedup at 720P while maintaining competitive video quality and similarity metrics.

📄 PDF Abstract BibTeX arXiv:2605.14513

Code (0)

등록된 구현이 없습니다.

Tasks

Video Generation

Similar Papers 제목 키워드 기반

REPA Works Until It Doesn't: Early-Stopped, Holistic Alignment Supercharges Diffusion Training

2025-05-22 · Ziqiao Wang, Wangbo Zhao, Yuhao Zhou, Zekai Li 외

Diffusion Transformers (DiTs) deliver state-of-the-art image quality, yet their training remains notoriously slow. A recent remedy -- representation alignment (REPA) that matches DiT hidden features to those of a non-gen…

Denoising

AsymRnR: Video Diffusion Transformers Acceleration with Asymmetric Reduction and Restoration

2024-12-16 · Wenhao Sun, Rong-Cheng Tu, Jingyi Liao, Zhao Jin 외

Video Diffusion Transformers (DiTs) have demonstrated significant potential for generating high-fidelity videos but are computationally intensive. Existing acceleration methods include distillation, which requires costly…

DenoisingToken Reduction

FasterCache: Training-Free Video Diffusion Model Acceleration with High Quality

2024-10-25 · Zhengyao Lv, Chenyang Si, Junhao Song, Zhenyu Yang 외

In this paper, we present \textbf{\textit{FasterCache}}, a novel training-free strategy designed to accelerate the inference of video diffusion models with high-quality generation. By analyzing existing cache-based metho…

Video Generation

LightCache: Memory-Efficient, Training-Free Acceleration for Video Generation

2025-10-06 · Yang Xiao, Gen Li, Kaiyuan Deng, Yushu Wu 외 arxiv

Training-free acceleration has emerged as an advanced research area in video generation based on diffusion models. The redundancy of latents in diffusion model inference provides a natural entry point for acceleration. I…

Video Generation

F3-Pruning: A Training-Free and Generalized Pruning Strategy towards Faster and Finer Text-to-Video Synthesis

2023-12-06 · Sitong Su, Jianzhi Liu, Lianli Gao, Jingkuan Song

Recently Text-to-Video (T2V) synthesis has undergone a breakthrough by training transformers or diffusion models on large-scale datasets. Nevertheless, inferring such large models incurs huge costs.Previous inference acc…