T2S: High-resolution Time Series Generation with Text-to-Series Diffusion Models

TL;DR

This paper introduces T2S, a diffusion-based framework for high-resolution, length-adaptive text-to-time series generation, addressing data scarcity and domain generalization issues with a new dataset and innovative model design.

Contribution

The work presents a novel, domain-agnostic diffusion model for text-to-time series generation that can produce sequences of arbitrary lengths, supported by a new large-scale dataset.

Findings

T2S achieves state-of-the-art results on 13 datasets across 12 domains.

The model effectively generates high-resolution, variable-length time series from textual descriptions.

The approach demonstrates strong generalization to diverse time series captions.

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.