{"slug": "prismflow-residual-dynamics-for-flow-matching-in-time-series-generation", "title": "PrismFlow: Residual Dynamics for Flow Matching in Time-Series Generation", "summary": "Researchers have developed PrismFlow, a new flow matching method that uses Koopman-inspired dynamical experts to generate high-quality time-series data. The approach addresses spectral distortion and poor mode coverage in standard flow matching by applying residual corrections through a confidence-aware Winner-Take-All objective. PrismFlow achieves state-of-the-art performance with a 15.6% gain in Context-FID and a 38.6% improvement in Discriminative Score, while remaining effective for forecasting and imputation tasks.", "body_md": "arXiv:2605.28867v1 Announce Type: new\nAbstract: Generating high-quality time-series data is challenging because real-world signals often exhibit multimodal patterns and multiscale dynamics, including oscillations and high-frequency variations. Flow Matching (FM) offers an efficient alternative to diffusion models, but practical implementations typically rely on a single finite-capacity global vector-field estimator. In such heterogeneous temporal distributions, distinct regimes may pass through nearby flow states while requiring incompatible conditional velocities. A monolithic estimator trained with the standard $\\ell_2$ velocity-matching objective may therefore learn an overly smoothed approximation of the local transport field. This estimator-level smoothing can attenuate branch-specific dynamics, leading to spectral distortion and poor mode coverage. To address this, we propose PrismFlow, a new FM method with Koopman-inspired dynamical experts. Each expert learns residual corrections in a latent space where local nonlinear temporal evolution can be approximated by linear transitions. We further propose a confidence-aware Winner-Take-All (WTA) objective that updates only the expert best aligned with each sample while masking gradients to the others, encouraging mode-specific specialization. During sampling, the selected expert adds a residual dynamical correction to the global transport field, preserving FM stability while recovering fine-grained and high-frequency temporal structures. Across various benchmarks, PrismFlow effectively mitigates the spectral contraction in standard FM and achieves state-of-the-art performance, with a 15.6% gain in Context-FID and a 38.6% improvement in Discriminative Score, while remaining robust in low-data settings and effective for forecasting and imputation.", "url": "https://wpnews.pro/news/prismflow-residual-dynamics-for-flow-matching-in-time-series-generation", "canonical_source": "https://arxiv.org/abs/2605.28867", "published_at": "2026-05-29 04:00:00+00:00", "updated_at": "2026-05-29 04:17:49.368076+00:00", "lang": "en", "topics": ["machine-learning", "generative-ai", "artificial-intelligence", "neural-networks", "ai-research"], "entities": ["PrismFlow", "Koopman"], "alternates": {"html": "https://wpnews.pro/news/prismflow-residual-dynamics-for-flow-matching-in-time-series-generation", "markdown": "https://wpnews.pro/news/prismflow-residual-dynamics-for-flow-matching-in-time-series-generation.md", "text": "https://wpnews.pro/news/prismflow-residual-dynamics-for-flow-matching-in-time-series-generation.txt", "jsonld": "https://wpnews.pro/news/prismflow-residual-dynamics-for-flow-matching-in-time-series-generation.jsonld"}}