T2S: High-resolution Time Series Generation with Text-to-Series Diffusion Models
Yunfeng Ge, Jiawei Li, Yiji Zhao, Haomin Wen, Zhao Li, Meikang Qiu, Hongyan Li, Ming Jin, Shirui Pan
TL;DR
This work tackles the challenges of text-to-time-series generation by introducing TSFragment-600K, a high-resolution fragment-level dataset, and T2S, a domain-agnostic diffusion-based framework that generates semantically aligned time series of arbitrary lengths. T2S fuses a length-adaptive VAE with a Flow Matching backbone and a Diffusion Transformer denoiser, and it leverages classifier-free guidance to balance conditional and unconditional generation. An interleaved training scheme across multiple lengths enables cross-length generalization without length-specific training, achieving state-of-the-art results across 13 datasets in 12 domains. The approach demonstrates strong performance on fragment-, point-, and instance-level captions, highlighting its practical impact for data-scarce, multimodal time-series applications in real-world settings.
Abstract
Text-to-Time Series generation holds significant potential to address challenges such as data sparsity, imbalance, and limited availability of multimodal time series datasets across domains. While diffusion models have achieved remarkable success in Text-to-X (e.g., vision and audio data) generation, their use in time series generation remains in its nascent stages. Existing approaches face two critical limitations: (1) the lack of systematic exploration of general-proposed time series captions, which are often domain-specific and struggle with generalization; and (2) the inability to generate time series of arbitrary lengths, limiting their applicability to real-world scenarios. In this work, we first categorize time series captions into three levels: point-level, fragment-level, and instance-level. Additionally, we introduce a new fragment-level dataset containing over 600,000 high-resolution time series-text pairs. Second, we propose Text-to-Series (T2S), a diffusion-based framework that bridges the gap between natural language and time series in a domain-agnostic manner. T2S employs a length-adaptive variational autoencoder to encode time series of varying lengths into consistent latent embeddings. On top of that, T2S effectively aligns textual representations with latent embeddings by utilizing Flow Matching and employing Diffusion Transformer as the denoiser. We train T2S in an interleaved paradigm across multiple lengths, allowing it to generate sequences of any desired length. Extensive evaluations demonstrate that T2S achieves state-of-the-art performance across 13 datasets spanning 12 domains.