VisionTS++: Cross-Modal Time Series Foundation Model with Continual Pre-trained Vision Backbones

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Shen, Lefei, Chen, Mouxiang, Liu, Xu, Fu, Han, Ren, Xiaoxue, Sun, Jianling, Li, Zhuo, Liu, Chenghao
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909833602531328
author Shen, Lefei
Chen, Mouxiang
Liu, Xu
Fu, Han
Ren, Xiaoxue
Sun, Jianling
Li, Zhuo
Liu, Chenghao
author_facet Shen, Lefei
Chen, Mouxiang
Liu, Xu
Fu, Han
Ren, Xiaoxue
Sun, Jianling
Li, Zhuo
Liu, Chenghao
contents Recent studies have indicated that vision models pre-trained on images can serve as time series foundation models (TSFMs) by reformulating time series forecasting (TSF) as image reconstruction. However, effective cross-modal transfer from vision to time series remains challenging due to three discrepancies: (1) the data-modality gap between structured, bounded image data and unbounded, heterogeneous time series; (2) the multivariate-forecasting gap between fixed RGB-three-channel vision models and time series with arbitrary numbers of variates; and (3) the probabilistic-forecasting gap between the deterministic outputs of vision models and the requirement for uncertainty-aware probabilistic predictions. To bridge these gaps, we propose VisonTS++, a TSFM based on continual pre-training of a vision model on large-scale time series. Our approach introduces three key innovations: (1) vision-model-based filtering to identify high-quality sequences to stabilize pre-training and mitigate modality gap; (2) colorized multivariate conversion, encoding multivariate series as multi-subfigure RGB images to enhance cross-variate modeling; (3) multi-quantile forecasting, using parallel reconstruction heads to generate quantile forecasts without parametric assumptions. Experiments show that VisionTS++ achieves state-of-the-art performance in both in-distribution and out-of-distribution forecasting, outperforming specialized TSFMs by 6%-44% in MSE reduction and ranking first in GIFT-Eval benchmark which comprises 23 datasets across 7 domains. Our work demonstrates that with appropriate adaptation, vision models can effectively generalize to TSF, thus advancing the pursuit of universal TSFMs. Code is available at https://github.com/HALF111/VisionTSpp.
format Preprint
id arxiv_https___arxiv_org_abs_2508_04379
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle VisionTS++: Cross-Modal Time Series Foundation Model with Continual Pre-trained Vision Backbones
Shen, Lefei
Chen, Mouxiang
Liu, Xu
Fu, Han
Ren, Xiaoxue
Sun, Jianling
Li, Zhuo
Liu, Chenghao
Computer Vision and Pattern Recognition
Machine Learning
Recent studies have indicated that vision models pre-trained on images can serve as time series foundation models (TSFMs) by reformulating time series forecasting (TSF) as image reconstruction. However, effective cross-modal transfer from vision to time series remains challenging due to three discrepancies: (1) the data-modality gap between structured, bounded image data and unbounded, heterogeneous time series; (2) the multivariate-forecasting gap between fixed RGB-three-channel vision models and time series with arbitrary numbers of variates; and (3) the probabilistic-forecasting gap between the deterministic outputs of vision models and the requirement for uncertainty-aware probabilistic predictions. To bridge these gaps, we propose VisonTS++, a TSFM based on continual pre-training of a vision model on large-scale time series. Our approach introduces three key innovations: (1) vision-model-based filtering to identify high-quality sequences to stabilize pre-training and mitigate modality gap; (2) colorized multivariate conversion, encoding multivariate series as multi-subfigure RGB images to enhance cross-variate modeling; (3) multi-quantile forecasting, using parallel reconstruction heads to generate quantile forecasts without parametric assumptions. Experiments show that VisionTS++ achieves state-of-the-art performance in both in-distribution and out-of-distribution forecasting, outperforming specialized TSFMs by 6%-44% in MSE reduction and ranking first in GIFT-Eval benchmark which comprises 23 datasets across 7 domains. Our work demonstrates that with appropriate adaptation, vision models can effectively generalize to TSF, thus advancing the pursuit of universal TSFMs. Code is available at https://github.com/HALF111/VisionTSpp.
title VisionTS++: Cross-Modal Time Series Foundation Model with Continual Pre-trained Vision Backbones
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2508.04379