FastSTI: A Fast Conditional Pseudo Numerical Diffusion Model for Spatio-temporal Traffic Data Imputation

TL;DR

This work proposes a fast conditional diffusion model for spatiotemporal traffic data imputation (FastSTI), and proposes the application of a high-order pseudo-numerical solver to speed up the process yet, obtain better performance.

Abstract

High-quality spatiotemporal traffic data is crucial for intelligent transportation systems (ITS) and their data-driven applications. Inevitably, the issue of missing data caused by various disturbances threatens the reliability of data acquisition. Recent studies of diffusion probability models have demonstrated the superiority of deep generative models in imputation tasks by precisely capturing the spatio-temporal correlation of traffic data. One drawback of diffusion models is their slow sampling/denoising process. In this work, we aim to accelerate the imputation process while retaining the performance. We propose a fast conditional diffusion model for spatiotemporal traffic data imputation (FastSTI). To speed up the process yet, obtain better performance, we propose the application of a high-order pseudo-numerical solver. Our method further revs the imputation by introducing a predefined alignment strategy of variance schedule during the sampling process. Evaluating FastSTI on two types of real-world traffic datasets (traffic speed and flow) with different missing data scenarios proves its ability to impute higher-quality samples in only six sampling steps, especially under high missing rates (60\% $\sim$ 90\%). The experimental results illustrate a speed-up of $\textbf{8.3} \times$ faster than the current state-of-the-art model while achieving better performance.

Figures (9)

• Figure 1: Proposed FastSTI Model Architecture. In FastSTI, we take observed values and geographic location information as input. Our approach uses linear interpolation and leverages the conditional feature prior module to model the prior spatiotemporal context. Afterwards, the prior feature weights are obtained and fed into the noise prediction module to help predict noise.
• Figure 2: Accelerated Imputation. FastSTI utilizes "schedule alignment" to estimate the denoised distribution, replacing multiple classical denoising steps, thereby accelerating inference without significant loss of accuracy.
• Figure 3: The Impact of Missing Rate on the Imputation Performance on METR-LA Dataset.
• Figure 4: The Impact of Missing Rate on the Imputation Performance on PEMS04 Dataset.
• Figure 5: Inference Times on Traffic Speed Datasets (METR-LA and PEMS-BAY).