FlexDiT: Dynamic Token Density Control for Diffusion Transformer

• URL: http://arxiv.org/abs/2412.06028v1
• Date: Sun, 08 Dec 2024 18:59:16 GMT
• Title: FlexDiT: Dynamic Token Density Control for Diffusion Transformer
• Authors: Shuning Chang, Pichao Wang, Jiasheng Tang, Yi Yang,
• Abstract summary: Diffusion Transformers (DiT) deliver impressive generative performance but face prohibitive computational demands.<n>We propose FlexDiT, a framework that dynamically adapts token density across both spatial and temporal dimensions.<n>Our experiments demonstrate FlexDiT's effectiveness, achieving a 55% reduction in FLOPs and a 175% improvement in inference speed.
• Score: 31.799640242972373
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Diffusion Transformers (DiT) deliver impressive generative performance but face prohibitive computational demands due to both the quadratic complexity of token-based self-attention and the need for extensive sampling steps. While recent research has focused on accelerating sampling, the structural inefficiencies of DiT remain underexplored. We propose FlexDiT, a framework that dynamically adapts token density across both spatial and temporal dimensions to achieve computational efficiency without compromising generation quality. Spatially, FlexDiT employs a three-segment architecture that allocates token density based on feature requirements at each layer: Poolingformer in the bottom layers for efficient global feature extraction, Sparse-Dense Token Modules (SDTM) in the middle layers to balance global context with local detail, and dense tokens in the top layers to refine high-frequency details. Temporally, FlexDiT dynamically modulates token density across denoising stages, progressively increasing token count as finer details emerge in later timesteps. This synergy between FlexDiT's spatially adaptive architecture and its temporal pruning strategy enables a unified framework that balances efficiency and fidelity throughout the generation process. Our experiments demonstrate FlexDiT's effectiveness, achieving a 55% reduction in FLOPs and a 175% improvement in inference speed on DiT-XL with only a 0.09 increase in FID score on 512$\times$512 ImageNet images, a 56% reduction in FLOPs across video generation datasets including FaceForensics, SkyTimelapse, UCF101, and Taichi-HD, and a 69% improvement in inference speed on PixArt-$\alpha$ on text-to-image generation task with a 0.24 FID score decrease. FlexDiT provides a scalable solution for high-quality diffusion-based generation compatible with further sampling optimization techniques.

Related papers

• FUSE: Label-Free Image-Event Joint Monocular Depth Estimation via Frequency-Decoupled Alignment and Degradation-Robust Fusion [63.87313550399871]
  Image-event joint depth estimation methods leverage complementary modalities for robust perception, yet face challenges in generalizability. We propose Self-supervised Transfer (PST) and FrequencyDe-coupled Fusion module (FreDF) PST establishes cross-modal knowledge transfer through latent space alignment with image foundation models. FreDF explicitly decouples high-frequency edge features from low-frequency structural components, resolving modality-specific frequency mismatches.
  arXiv  Detail & Related papers  (2025-03-25T15:04:53Z)
• Training-free and Adaptive Sparse Attention for Efficient Long Video Generation [31.615453637053793]
  generating high-fidelity long videos with Diffusion Transformers (DiTs) is often hindered by significant latency. We propose AdaSpa, the first Dynamic Pattern and Online Precise Search sparse attention method. AdaSpa is implemented as an adaptive, plug-and-play solution and can be integrated seamlessly with existing DiTs.
  arXiv  Detail & Related papers  (2025-02-28T14:11:20Z)
• ContextFormer: Redefining Efficiency in Semantic Segmentation [48.81126061219231]
  Convolutional methods, although capturing local dependencies well, struggle with long-range relationships. Vision Transformers (ViTs) excel in global context capture but are hindered by high computational demands. We propose ContextFormer, a hybrid framework leveraging the strengths of CNNs and ViTs in the bottleneck to balance efficiency, accuracy, and robustness for real-time semantic segmentation.
  arXiv  Detail & Related papers  (2025-01-31T16:11:04Z)
• Layer- and Timestep-Adaptive Differentiable Token Compression Ratios for Efficient Diffusion Transformers [55.87192133758051]
  Diffusion Transformers (DiTs) have achieved state-of-the-art (SOTA) image generation quality but suffer from high latency and memory inefficiency. We propose DiffCR, a dynamic DiT inference framework with differentiable compression ratios.
  arXiv  Detail & Related papers  (2024-12-22T02:04:17Z)
• Dynamic Diffusion Transformer [67.13876021157887]
  Diffusion Transformer (DiT) has demonstrated superior performance but suffers from substantial computational costs. We propose Dynamic Diffusion Transformer (DyDiT), an architecture that dynamically adjusts its computation along both timestep and spatial dimensions during generation. With 3% additional fine-tuning, our method reduces the FLOPs of DiT-XL by 51%, accelerates generation by 1.73, and achieves a competitive FID score of 2.07 on ImageNet.
  arXiv  Detail & Related papers  (2024-10-04T14:14:28Z)
• Flexiffusion: Segment-wise Neural Architecture Search for Flexible Denoising Schedule [50.260693393896716]
  Diffusion models are cutting-edge generative models adept at producing diverse, high-quality images. Recent techniques have been employed to automatically search for faster generation processes. We introduce Flexiffusion, a novel training-free NAS paradigm designed to accelerate diffusion models.
  arXiv  Detail & Related papers  (2024-09-26T06:28:05Z)
• LeRF: Learning Resampling Function for Adaptive and Efficient Image Interpolation [64.34935748707673]
  Recent deep neural networks (DNNs) have made impressive progress in performance by introducing learned data priors. We propose a novel method of Learning Resampling (termed LeRF) which takes advantage of both the structural priors learned by DNNs and the locally continuous assumption. LeRF assigns spatially varying resampling functions to input image pixels and learns to predict the shapes of these resampling functions with a neural network.
  arXiv  Detail & Related papers  (2024-07-13T16:09:45Z)
• The Surprising Effectiveness of Skip-Tuning in Diffusion Sampling [78.6155095947769]
  Skip-Tuning is a simple yet surprisingly effective training-free tuning method on the skip connections. Our method can achieve 100% FID improvement for pretrained EDM on ImageNet 64 with only 19 NFEs (1.75) While Skip-Tuning increases the score-matching losses in the pixel space, the losses in the feature space are reduced.
  arXiv  Detail & Related papers  (2024-02-23T08:05:23Z)
• Dynamic PlenOctree for Adaptive Sampling Refinement in Explicit NeRF [6.135925201075925]
  We propose the dynamic PlenOctree DOT, which adaptively refines the sample distribution to adjust to changing scene complexity. Compared with POT, our DOT outperforms it by enhancing visual quality, reducing over $55.15$/$68.84%$ parameters, and providing 1.7/1.9 times FPS for NeRF-synthetic and Tanks $&$ Temples, respectively.
  arXiv  Detail & Related papers  (2023-07-28T06:21:42Z)
• Efficient Context Integration through Factorized Pyramidal Learning for Ultra-Lightweight Semantic Segmentation [1.0499611180329804]
  We propose a novel Factorized Pyramidal Learning (FPL) module to aggregate rich contextual information in an efficient manner. We decompose the spatial pyramid into two stages which enables a simple and efficient feature fusion within the module to solve the notorious checkerboard effect. Based on the FPL module and FIR unit, we propose an ultra-lightweight real-time network, called FPLNet, which achieves state-of-the-art accuracy-efficiency trade-off.
  arXiv  Detail & Related papers  (2023-02-23T05:34:51Z)
• DSVT: Dynamic Sparse Voxel Transformer with Rotated Sets [95.84755169585492]
  We present Dynamic Sparse Voxel Transformer (DSVT), a single-stride window-based voxel Transformer backbone for outdoor 3D perception. Our model achieves state-of-the-art performance with a broad range of 3D perception tasks.
  arXiv  Detail & Related papers  (2023-01-15T09:31:58Z)
• Tutel: Adaptive Mixture-of-Experts at Scale [20.036168971435306]
  Sparsely-gated mixture-of-experts (MoE) has been widely adopted to scale deep learning models to trillion-plus parameters with fixed computational cost. We present Flex, a highly scalable stack design and implementation for MoE with dynamically adaptive parallelism and pipelining. Our evaluation shows that Flex efficiently and effectively runs a real-world MoE-based model named SwinV2-MoE, built upon Swin Transformer V2, a state-of-the-art computer vision architecture.
  arXiv  Detail & Related papers  (2022-06-07T15:20:20Z)

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.