Retrieving and Refining Winning Noise Tickets for Diffusion-Based Motion Generation

• URL: http://arxiv.org/abs/2607.06843v1
• Date: Tue, 07 Jul 2026 22:32:10 GMT
• Title: Retrieving and Refining Winning Noise Tickets for Diffusion-Based Motion Generation
• Abstract summary: Diffusion-based text-to-motion models synthesize realistic human motions but often exhibit semantic drift from the input text.<n>We propose WInning Noise Retrieval and Optimization (WINRO), a framework that improves text-motion alignment by selecting and refining such tickets before diffusion sampling.<n>WINRO consistently improves text-motion fidelity across different base models, MDM and MotionLCM, on HumanML3D without retraining.
• Score: 32.824608968455294
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Diffusion-based text-to-motion models synthesize realistic human motions but often exhibit semantic drift from the input text. Motion is inherently temporal, especially in compositional and long-duration sequences that require semantic consistency across multiple action segments and smooth kinematic transitions throughout the trajectory. We posit that the initial noise is central to this consistency: within the Gaussian noise space, certain instances, i.e. winning noise tickets, carry latent structure that biases denoising toward particular motion semantics, even under null prompts. We propose WInning Noise Retrieval and Optimization (WINRO), a training-free, model-agnostic framework that improves text-motion alignment by selecting and refining such tickets before diffusion sampling. WINRO maps random noises to motion features generated under null prompts, retrieves the best-aligned noise for a given text, and refines it via a KL-regularized objective that reduces the residual semantic gap while preserving the Gaussian prior. An optional LoRA-based adapter amortizes this refinement into a single forward pass. WINRO consistently improves text-motion fidelity across different base models, MDM and MotionLCM, on HumanML3D without retraining, improves temporal robustness on the MTT benchmark, and generalizes to applications such as motion stylization and spatial constraint satisfaction.

Related papers

• Beyond Noisy Signals: Dual-Level Denoising for Multi-modal Sequential Recommendation [6.902443075506667]
  We propose textbfDual-level textbfDenoising textbfMulti-modal textbfSequential textbfRecommendation framework.
  arXiv  Detail & Related papers  (2026-07-21T07:06:06Z)
• RhymeFlow: Training-Free Acceleration for Video Generation with Asynchronous Denoising Flow Scheduling [51.279397568734424]
  Video generation models based on Diffusion Transformers (DiTs) have achieved remarkable performance in video synthesis.<n>DiTs suffer from high inference latency and computational costs due to the quadratic complexity of 3D attention.<n>We introduce bftextRhymeFlow, a training-free framework that decouples the denoising trajectories of different frames.
  arXiv  Detail & Related papers  (2026-06-04T15:49:37Z)
• Motion-Adaptive Temporal Attention for Lightweight Video Generation with Stable Diffusion [1.9349092719498848]
  We present a motion-adaptive temporal attention mechanism for parameter-efficient video generation built upon frozen Stable Diffusion models.<n>We inject lightweight temporal attention modules into all UNet transformer blocks via a cascaded strategy.<n>Our ablation studies reveal a clear trade-off between noise correlation and motion amplitude, providing a practical inference-time control for diverse generation behaviors.
  arXiv  Detail & Related papers  (2026-03-18T06:20:57Z)
• MotionCFG: Boosting Motion Dynamics via Stochastic Concept Perturbation [60.48914865049489]
  MotionCFG is a framework that enhances motion dynamics by contrasting a target concept with its noise-blured counterparts.<n>Unlike explicit negations, this approach facilitates implicit hard negative guidance without shifting the global semantic identity.<n> MotionCFG consistently improves motion dynamics across state-of-the-art T2V frameworks.
  arXiv  Detail & Related papers  (2026-03-14T18:39:37Z)
• DEMO: Disentangled Motion Latent Flow Matching for Fine-Grained Controllable Talking Portrait Synthesis [15.304037069236536]
  DEMO is a flow-matching generative framework for audio-driven talking-head video synthesis.<n>It delivers disentangled, high-fidelity control of lip motion, head pose, and eye gaze.
  arXiv  Detail & Related papers  (2025-10-12T15:10:33Z)
• MotionFlux: Efficient Text-Guided Motion Generation through Rectified Flow Matching and Preference Alignment [38.42799902378583]
  Motion generation is essential for animating virtual characters and embodied agents.<n>TAPO and MotionFLUX form a unified system that outperforms state-of-the-art approaches in both semantic consistency and motion quality.
  arXiv  Detail & Related papers  (2025-08-27T02:45:09Z)
• DiTVR: Zero-Shot Diffusion Transformer for Video Restoration [48.97196894658511]
  DiTVR is a zero shot video restoration framework that couples a diffusion transformer with trajectory aware attention and a flow consistent sampler.<n>Our attention mechanism aligns tokens along optical flow trajectories, with particular emphasis on vital layers that exhibit the highest sensitivity to temporal dynamics.<n>The flow guided sampler injects data consistency only into low-frequency bands, preserving high frequency priors while accelerating cache.
  arXiv  Detail & Related papers  (2025-08-11T09:54:45Z)
• WaMo: Wavelet-Enhanced Multi-Frequency Trajectory Analysis for Fine-Grained Text-Motion Retrieval [7.349030413222046]
  Text-Motion Retrieval aims to retrieve 3D motion sequences semantically relevant to text descriptions.<n>We propose WaMo, a novel wavelet-based multi-frequency feature extraction framework.<n>WaMo captures part-specific and time-varying motion details across multiple resolutions on body joints.
  arXiv  Detail & Related papers  (2025-08-05T11:44:26Z)
• FTMoMamba: Motion Generation with Frequency and Text State Space Models [53.60865359814126]
  We propose a novel diffusion-based FTMoMamba framework equipped with a Frequency State Space Model and a Text State Space Model. To learn fine-grained representation, FreqSSM decomposes sequences into low-frequency and high-frequency components. To ensure the consistency between text and motion, TextSSM encodes text features at the sentence level.
  arXiv  Detail & Related papers  (2024-11-26T15:48:12Z)
• DiffusionPhase: Motion Diffusion in Frequency Domain [69.811762407278]
  We introduce a learning-based method for generating high-quality human motion sequences from text descriptions. Existing techniques struggle with motion diversity and smooth transitions in generating arbitrary-length motion sequences. We develop a network encoder that converts the motion space into a compact yet expressive parameterized phase space.
  arXiv  Detail & Related papers  (2023-12-07T04:39:22Z)
• ProgressiveMotionSeg: Mutually Reinforced Framework for Event-Based Motion Segmentation [101.19290845597918]
  This paper presents a Motion Estimation (ME) module and an Event Denoising (ED) module jointly optimized in a mutually reinforced manner. Taking temporal correlation as guidance, ED module calculates the confidence that each event belongs to real activity events, and transmits it to ME module to update energy function of motion segmentation for noise suppression.
  arXiv  Detail & Related papers  (2022-03-22T13:40:26Z)

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.