VMoBA: Mixture-of-Block Attention for Video Diffusion Models

• URL: http://arxiv.org/abs/2506.23858v1
• Date: Mon, 30 Jun 2025 13:52:31 GMT
• Title: VMoBA: Mixture-of-Block Attention for Video Diffusion Models
• Abstract summary: This paper introduces Video Mixture of Block Attention (VMoBA), a novel attention mechanism specifically adapted for Video Diffusion Models (VDMs)<n>Motivated by an in-depth analysis of attention patterns within pre-trained video transformers, VMoBA enhances the original MoBA framework with three key modifications.<n>Experiments demonstrate that VMoBA significantly accelerates the training of VDMs on longer sequences, achieving 2.92x FLOPs and 1.48x latency speedup.
• Score: 29.183614108287276
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The quadratic complexity of full attention mechanisms poses a significant bottleneck for Video Diffusion Models (VDMs) aiming to generate long-duration, high-resolution videos. While various sparse attention methods have been proposed, many are designed as training-free inference accelerators or do not optimally capture the unique spatio-temporal characteristics inherent in video data when trained natively. This paper introduces Video Mixture of Block Attention (VMoBA), a novel sparse attention mechanism specifically adapted for VDMs. Motivated by an in-depth analysis of attention patterns within pre-trained video transformers, which revealed strong spatio-temporal locality, varying query importance, and head-specific concentration levels, VMoBA enhances the original MoBA framework with three key modifications: (1) a layer-wise recurrent block partition scheme (1D-2D-3D) to dynamically adapt to diverse spatio-temporal attention patterns and improve efficiency; (2) global block selection to prioritize the most salient query-key block interactions across an entire attention head; and (3) threshold-based block selection to dynamically determine the number of attended blocks based on their cumulative similarity. Extensive experiments demonstrate that VMoBA significantly accelerates the training of VDMs on longer sequences, achieving 2.92x FLOPs and 1.48x latency speedup, while attaining comparable or even superior generation quality to full attention. Furthermore, VMoBA exhibits competitive performance in training-free inference, offering 2.40x FLOPs and 1.35x latency speedup for high-res video generation.

Related papers

• Chimera: Designing and Chinchilla-Scaling Hybrid Visual Diffusion Transformers [54.01042826416009]
  Visual generation requires high-resolution images, long videos, and multimodal context.<n>We introduce Chimera, a hybrid visual diffusion backbone with a principled scaling recipe.<n>We show that the dense backbone is 1.7x as compute-efficient as a matched full-attention Wan2.1 2B baseline.
  arXiv  Detail & Related papers  (2026-07-30T17:58:02Z)
• Kwai Keye-VL-2.0 Technical Report [53.82434681649277]
  Keye-VL-2.0 is a multimodal foundation model designed to advance long-video understanding and agentic intelligence.<n>It is the first to adapt DeepSeek Sparse Attention (DSA) to GQA-based multimodal architectures.<n>Cross-Modal Multi-Teacher On-Policy Distillation (MOPD) paired with Context-RL and Video-RL overcomes the algorithmic dilemma of catastrophic forgetting.
  arXiv  Detail & Related papers  (2026-06-09T09:58:08Z)
• HASTE: Training-Free Video Diffusion Acceleration via Head-Wise Adaptive Sparse Attention [68.95533683996236]
  Training-free sparse attention is attractive because it accelerates pretrained models without retraining.<n>Existing online top-$p$ sparse attention still spends non-negligible cost on mask prediction.<n>We show that these two overlooked factors limit the practical speed-quality trade-off of training-free sparse attention in Video DiTs.
  arXiv  Detail & Related papers  (2026-05-14T07:57:55Z)
• SDAR-VL: Stable and Efficient Block-wise Diffusion for Vision-Language Understanding [25.2227348401136]
  Block-wise discrete diffusion offers an attractive balance between parallel generation and causal dependency modeling.<n>We present textbfSDAR-VL, the first systematic application of block-wise discrete diffusion to large-scale vision-language understanding.<n>We show that SDAR-VL consistently improves emphtraining efficiency, emphconvergence stability, and emphtask performance over conventional block diffusion.
  arXiv  Detail & Related papers  (2025-12-16T04:12:52Z)
• USV: Unified Sparsification for Accelerating Video Diffusion Models [11.011602744993942]
  Unified Sparsification for Video diffusion models is an end-to-end trainable framework.<n>It orchestrates sparsification across both the model's internal computation and its sampling process.<n>It achieves up to 83.3% speedup in the denoising process and 22.7% end-to-end acceleration, while maintaining high visual fidelity.
  arXiv  Detail & Related papers  (2025-12-05T14:40:06Z)
• Bidirectional Sparse Attention for Faster Video Diffusion Training [14.523882232476092]
  Video diffusion Transformer (DiT) models excel in generative quality but hit major computational bottlenecks when producing high-resolution, long-duration videos.<n>We propose a Bidirectional Sparse Attention (BSA) framework for faster video DiT training, the first to dynamically sparsify both Queries and Key-Value pairs within 3D full attention.<n>BSA significantly accelerates DiT training across long sequences, reducing FLOPs by up to 20x and achieving 17.79x faster attention training, while preserving or even surpassing the generative quality of full attention.
  arXiv  Detail & Related papers  (2025-09-01T03:16:52Z)
• MoCa: Modality-aware Continual Pre-training Makes Better Bidirectional Multimodal Embeddings [75.0617088717528]
  MoCa is a framework for transforming pre-trained VLM backbones into effective bidirectional embedding models.<n>MoCa consistently improves performance across MMEB and ViDoRe-v2 benchmarks, achieving new state-of-the-art results.
  arXiv  Detail & Related papers  (2025-06-29T06:41:00Z)
• VORTA: Efficient Video Diffusion via Routing Sparse Attention [45.269274789183974]
  Video Diffusion Transformers (VDiTs) have achieved remarkable progress in high-quality video generation, but remain computationally expensive.<n>We propose VORTA, an acceleration framework with two novel components.<n>It achieves a $1.76times$ end-to-end speedup without quality loss on VBench.
  arXiv  Detail & Related papers  (2025-05-24T17:46:47Z)
• Training-Free Efficient Video Generation via Dynamic Token Carving [54.52061549312799]
  Jenga is an inference pipeline that combines dynamic attention carving with progressive resolution generation.<n>As a plug-and-play solution, Jenga enables practical, high-quality video generation on modern hardware.
  arXiv  Detail & Related papers  (2025-05-22T16:21:32Z)
• ViaRL: Adaptive Temporal Grounding via Visual Iterated Amplification Reinforcement Learning [68.76048244253582]
  We introduce ViaRL, the first framework to leverage rule-based reinforcement learning (RL) for optimizing frame selection in video understanding.<n>ViaRL utilizes the answer accuracy of a downstream model as a reward signal to train a frame selector through trial-and-error.<n>ViaRL consistently delivers superior temporal grounding performance and robust generalization across diverse video understanding tasks.
  arXiv  Detail & Related papers  (2025-05-21T12:29:40Z)
• VSA: Faster Video Diffusion with Trainable Sparse Attention [21.593548582058403]
  Scaling video diffusion transformers (DiTs) is limited by their quadratic 3D attention, even though most of the attention mass concentrates on a small subset of positions.<n>We turn this observation into VSA, a trainable, hardware-efficient sparse attention that replaces full attention at emphboth training and inference.
  arXiv  Detail & Related papers  (2025-05-19T17:30:13Z)
• DiVE: Efficient Multi-View Driving Scenes Generation Based on Video Diffusion Transformer [56.98400572837792]
  DiVE produces high-fidelity, temporally coherent, and cross-view consistent multi-view videos.<n>These innovations collectively achieve a 2.62x speedup with minimal quality degradation.
  arXiv  Detail & Related papers  (2025-04-28T09:20:50Z)
• Towards Scalable Modeling of Compressed Videos for Efficient Action Recognition [6.168286187549952]
  We propose a hybrid end-to-end framework that factorizes learning across three key concepts to reduce inference cost by $330times$ versus prior art.<n> Experiments show that our method results in a lightweight architecture achieving state-of-the-art video recognition performance.
  arXiv  Detail & Related papers  (2025-03-17T21:13:48Z)
• DSV: Exploiting Dynamic Sparsity to Accelerate Large-Scale Video DiT Training [85.04885553561164]
  Diffusion Transformers (DiTs) have shown remarkable performance in generating high-quality videos.<n>DiTs can consume up to 95% of processing time and demand specialized context parallelism.<n>This paper introduces DSV to accelerate video DiT training by leveraging the dynamic attention sparsity we empirically observe.
  arXiv  Detail & Related papers  (2025-02-11T14:39:59Z)
• Vchitect-2.0: Parallel Transformer for Scaling Up Video Diffusion Models [89.79067761383855]
  Vchitect-2.0 is a parallel transformer architecture designed to scale up video diffusion models for large-scale text-to-video generation.<n>By introducing a novel Multimodal Diffusion Block, our approach achieves consistent alignment between text descriptions and generated video frames.<n>To overcome memory and computational bottlenecks, we propose a Memory-efficient Training framework.
  arXiv  Detail & Related papers  (2025-01-14T21:53:11Z)
• DiTCtrl: Exploring Attention Control in Multi-Modal Diffusion Transformer for Tuning-Free Multi-Prompt Longer Video Generation [54.30327187663316]
  DiTCtrl is a training-free multi-prompt video generation method under MM-DiT architectures for the first time.<n>We analyze MM-DiT's attention mechanism, finding that the 3D full attention behaves similarly to that of the cross/self-attention blocks in the UNet-like diffusion models.<n>Based on our careful design, the video generated by DiTCtrl achieves smooth transitions and consistent object motion given multiple sequential prompts.
  arXiv  Detail & Related papers  (2024-12-24T18:51:19Z)
• Multimodal Attention-Enhanced Feature Fusion-based Weekly Supervised Anomaly Violence Detection [1.9223495770071632]
  This system uses three feature streams: RGB video, optical flow, and audio signals, where each stream extracts complementary spatial and temporal features. The system significantly improves anomaly detection accuracy and robustness across three datasets.
  arXiv  Detail & Related papers  (2024-09-17T14:17:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.