VisionTS: Visual Masked Autoencoders Are Free-Lunch Zero-Shot Time Series Forecasters

TL;DR

This paper introduces VisionTS, a novel approach that leverages pre-trained visual masked autoencoders on natural images to perform zero-shot and fine-tuned time series forecasting, achieving state-of-the-art results without domain-specific training.

Contribution

It demonstrates that visual masked autoencoders can be directly applied to time series forecasting by reformulating the task as image reconstruction, bridging the gap between vision models and time series tasks.

Findings

VisionTS outperforms existing TSF foundation models in zero-shot settings.

Fine-tuning VisionTS for one epoch further improves forecasting accuracy.

Intrinsic similarities between images and time series suggest potential for cross-modality models.

Abstract

Foundation models have emerged as a promising approach in time series forecasting (TSF). Existing approaches either repurpose large language models (LLMs) or build large-scale time series datasets to develop TSF foundation models for universal forecasting. However, these methods face challenges due to the severe cross-domain gap or in-domain heterogeneity. This paper explores a new road to building a TSF foundation model from rich, high-quality natural images. Our key insight is that a visual masked autoencoder, pre-trained on the ImageNet dataset, can naturally be a numeric series forecaster. By reformulating TSF as an image reconstruction task, we bridge the gap between image pre-training and TSF downstream tasks. Surprisingly, without further adaptation in the time series domain, the proposed VisionTS could achieve better zero-shot forecast performance than existing TSF foundation models. With fine-tuning for one epoch, VisionTS could further improve the forecasting and achieve state-of-the-art performance in most cases. Extensive experiments reveal intrinsic similarities between images and real-world time series, suggesting that visual models may offer a "free lunch" for TSF and highlight the potential for future cross-modality research. Our code is publicly available at https://github.com/Keytoyze/VisionTS.