TimeOmni-VL: Unified Models for Time Series Understanding and Generation

TL;DR

TimeOmni-VL is a pioneering unified framework that combines high-fidelity time series understanding and generation by bridging the gap between numerical and semantic modeling through innovative bidirectional mappings and understanding-guided generation.

Contribution

It introduces a novel vision-centric model that unifies time series understanding and generation with near-lossless image conversions and a new dataset for comprehensive evaluation.

Findings

Enhanced semantic understanding of time series data.

Improved numerical precision in generated time series.

Established a new benchmark for multimodal time series modeling.

Abstract

Recent time series modeling faces a sharp divide between numerical generation and semantic understanding, with research showing that generation models often rely on superficial pattern matching, while understanding-oriented models struggle with high-fidelity numerical output. Although unified multimodal models (UMMs) have bridged this gap in vision, their potential for time series remains untapped. We propose TimeOmni-VL, the first vision-centric framework that unifies time series understanding and generation through two key innovations: (1) Fidelity-preserving bidirectional mapping between time series and images (Bi-TSI), which advances Time Series-to-Image (TS2I) and Image-to-Time Series (I2TS) conversions to ensure near-lossless transformations. (2) Understanding-guided generation. We introduce TSUMM-Suite, a novel dataset consists of six understanding tasks rooted in time series analytics that are coupled with two generation tasks. With a calibrated Chain-of-Thought, TimeOmni-VL is the first to leverage time series understanding as an explicit control signal for high-fidelity generation. Experiments confirm that this unified approach significantly improves both semantic understanding and numerical precision, establishing a new frontier for multimodal time series modeling.