FlowTS: Time Series Generation via Rectified Flow

TL;DR

FlowTS introduces an efficient ODE-based time series generation model using rectified flow and geodesic paths, significantly improving speed and performance over diffusion models, with versatile conditional and unconditional capabilities.

Contribution

The paper proposes FlowTS, a novel rectified flow model for time series generation that achieves computational efficiency and high-quality results through exact linear trajectories and adaptive sampling.

Findings

Achieves state-of-the-art performance on Stock and ETTh datasets.

Significantly reduces computation compared to diffusion-based models.

Outperforms previous methods in solar forecasting and MuJoCo imputation.

Abstract

Diffusion-based models have significant achievements in time series generation but suffer from inefficient computation: solving high-dimensional ODEs/SDEs via iterative numerical solvers demands hundreds to thousands of drift function evaluations per sample, incurring prohibitive costs. To resolve this, we propose FlowTS, an ODE-based model that leverages rectified flow with straight-line transport in probability space. By learning geodesic paths between distributions, FlowTS achieves computational efficiency through exact linear trajectory simulation, accelerating training and generation while improving performances. We further introduce an adaptive sampling strategy inspired by the exploration-exploitation trade-off, balancing noise adaptation and precision. Notably, FlowTS enables seamless adaptation from unconditional to conditional generation without retraining, ensuring efficient real-world deployment. Also, to enhance generation authenticity, FlowTS integrates trend and seasonality decomposition, attention registers (for global context aggregation), and Rotary Position Embedding (RoPE) (for position information). For unconditional setting, extensive experiments demonstrate that FlowTS achieves state-of-the-art performance, with context FID scores of 0.019 and 0.011 on Stock and ETTh datasets (prev. best: 0.067, 0.061). For conditional setting, we have achieved superior performance in solar forecasting (MSE 213, prev. best: 375) and MuJoCo imputation tasks (MSE 7e-5, prev. best 2.7e-4). The code is available at https://github.com/UNITES-Lab/FlowTS.