ORBIS: Output-Guided Token Reduction with Distribution-Aware Matching for Video Diffusion Acceleration

• URL: http://arxiv.org/abs/2605.22015v1
• Date: Thu, 21 May 2026 05:23:21 GMT
• Title: ORBIS: Output-Guided Token Reduction with Distribution-Aware Matching for Video Diffusion Acceleration
• Abstract summary: Diffusion Transformer (DiT) has emerged as a powerful model architecture for generating high-quality images and videos.<n>We propose ORBIS, an SW-HW co-designed accelerator for video DiT.<n>We show that ORBIS achieves about 2x higher token reduction ratio than the state-of-the-art approach, AsymRnR.
• Score: 2.918426765142262
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Diffusion Transformer (DiT) has emerged as a powerful model architecture for generating high-quality images and videos. In the case of video DiT, 3D Spatio-Temporal Attention increases token length in proportion to the number of frames, sharply increasing computational cost. Token reduction methods mitigate this cost by exploiting spatial redundancy, but existing approaches rely on inaccurate similarity estimates and lightweight matching algorithms, resulting in poor matching quality and only marginal acceleration. To overcome these limitations, we propose ORBIS, an SW-HW co-designed accelerator for video DiT. ORBIS leverages the output activation from the previous timestep to obtain more accurate inter-token similarity, substantially improving matching quality and enabling a higher token reduction ratio. We further introduce a Distribution-Aware Token Matching (DATM) algorithm that captures global token distribution and explicitly minimizes token-pair loss for additional gains. To fully hide DATM latency, we design specialized, deeply pipelined hardware and minimize its hardware cost through quantization, occupying only 2.4% of total area with negligible accuracy loss. Extensive experiments show that ORBIS achieves about 2x higher token reduction ratio than the state-of-the-art approach, AsymRnR, while delivering up to 4.5x speedup and 79.3% energy reduction compared to an NVIDIA A100 GPU.

Related papers

• Kaleido: Algorithm-Hardware Co-Design for Video Diffusion Transformers by Exploiting Latent Space Correlations [47.69636045010864]
  Video diffusion transformers (vDiTs) generate high quality video but introduce extremely high compute cost.<n>Existing acceleration approaches largely inherit attention techniques from sparse models which fail to consider unique video data.<n>This paper presents Kale Kale, an algorithm hardware codesign that accelerates all operations in vDiTs by exploiting channel-wisetemporal correlations in latent space.<n> Evaluations across three mainstream vDiT models show that Kale Kale achieves 5.9x speedup and 16.0x energy savings over state of the art accelerators.
  arXiv  Detail & Related papers  (2026-07-15T12:31:19Z)
• BWTA: Accurate and Efficient Binarized Transformer by Algorithm-Hardware Co-design [71.97035034203275]
  We analyze zero-point distortion in binarization and propose a Binary Weights & Ternary Activations (BWTA) quantization scheme.<n>For training, we propose Smooth Multi-Stage Quantization, combining a Levelwise Degradation Strategy and a MagnitudeAlignment Projection Factor.<n>Experiments show that BWTA approaches full-precision performance for Transformer-based models, with an average 3.5% drop on GLUE and less than 2% drop on five tasks.
  arXiv  Detail & Related papers  (2026-04-05T04:25:07Z)
• QuantSparse: Comprehensively Compressing Video Diffusion Transformer with Model Quantization and Attention Sparsification [67.15451442018258]
  Diffusion transformers exhibit remarkable video generation capability, yet their prohibitive computational and memory costs hinder practical deployment.<n>Model quantization and attention sparsification are two promising directions for compression, but each alone suffers severe performance degradation under aggressive compression.<n>We propose textbfQuantSparse, a unified framework that integrates model quantization with attention sparsification.
  arXiv  Detail & Related papers  (2025-09-28T06:49:44Z)
• ToMA: Token Merge with Attention for Diffusion Models [8.079656935981193]
  Diffusion models excel in high-fidelity image generation but face scalability limits due to transformers' quadratic attention complexity.<n>We propose Token Merge with Attention (ToMA), an off-the-shelf method that negates token reduction for GPU-aligned efficiency.<n>ToMA reduces SDXL/Flux generation latency by 24%/23%, respectively (with DINO $Delta 0.07$), outperforming prior methods.
  arXiv  Detail & Related papers  (2025-09-13T17:35:00Z)
• Astraea: A Token-wise Acceleration Framework for Video Diffusion Transformers [29.130090574300635]
  Video diffusion transformers (vDiTs) have made tremendous progress in text-to-video generation, but their compute demands pose a major challenge for practical deployment.<n>We introduce Astraea, a framework that searches for near-optimal configurations for vDiT-based video generation under a performance target.
  arXiv  Detail & Related papers  (2025-06-05T14:41:38Z)
• DDT: Decoupled Diffusion Transformer [51.84206763079382]
  Diffusion transformers encode noisy inputs to extract semantic component and decode higher frequency with identical modules.<n>textbfcolorddtDecoupled textbfcolorddtTransformer(textbfcolorddtDDT)<n>textbfcolorddtTransformer(textbfcolorddtDDT)<n>textbfcolorddtTransformer(textbfcolorddtDDT)
  arXiv  Detail & Related papers  (2025-04-08T07:17:45Z)
• Speedy MASt3R [68.47052557089631]
  MASt3R redefines image matching as a 3D task by leveraging DUSt3R and introducing a fast reciprocal matching scheme.<n>Fast MASt3R achieves a 54% reduction in inference time (198 ms to 91 ms per image pair) without sacrificing accuracy.<n>This advancement enables real-time 3D understanding, benefiting applications like mixed reality navigation and large-scale 3D scene reconstruction.
  arXiv  Detail & Related papers  (2025-03-13T03:56:22Z)
• Accelerating Error Correction Code Transformers [56.75773430667148]
  We introduce a novel acceleration method for transformer-based decoders. We achieve a 90% compression ratio and reduce arithmetic operation energy consumption by at least 224 times on modern hardware.
  arXiv  Detail & Related papers  (2024-10-08T11:07:55Z)
• Motion Guided Token Compression for Efficient Masked Video Modeling [7.548789718676737]
  This paper showcases the enhanced performance achieved through an escalation in the frames per second (FPS) rate. We also present a novel approach, Motion Guided Token Compression (MGTC), to empower Transformer models to utilize a smaller yet more representative set of tokens for comprehensive video representation. Our experiments, conducted on widely examined video recognition datasets, Kinetics-400, UCF101 and HMDB51, demonstrate that elevating the FPS rate results in a significant top-1 accuracy score improvement of over 1.6, 1.6 and 4.0.
  arXiv  Detail & Related papers  (2024-01-10T07:49:23Z)

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.