Time series forecasting is fundamental to diverse applications, with recent approaches leverage large vision models (LVMs) to capture temporal patterns through visual representations. We reveal that while vision models enhance forecasting performance, 99% of their parameters are unnecessary for time series tasks. Through cross-modal analysis, we find that time series align with low-level textural features but not high-level semantics, which can impair forecasting accuracy. We propose OccamVTS, a knowledge distillation framework that extracts only the essential 1% of predictive information from LVMs into lightweight networks. Using pre-trained LVMs as privileged teachers, OccamVTS employs pyramid-style feature alignment combined with correlation and feature distillation to transfer beneficial patterns while filtering out semantic noise. Counterintuitively, this aggressive parameter reduction improves accuracy by eliminating overfitting to irrelevant visual features while preserving essential temporal patterns. Extensive experiments across multiple benchmark datasets demonstrate that OccamVTS consistently achieves state-of-the-art performance with only 1% of the original parameters, particularly excelling in few-shot and zero-shot scenarios.

Traffic prediction serves as a cornerstone of modern intelligent transportation systems and the critical task of spatio-temporal forecasting. Although advanced Spatio-temporal Graph Neural Networks (STGNNs) and pre-trained models have made significant progress in traffic prediction, two critical challenges persist: (i) limited contextual capacity when handling complex spatio-temporal dependencies, and (ii) low predictability at fine-grained spatio-temporal points caused by heterogeneous patterns. Inspired by Retrieval-Augmented Generation (RAG), we propose RAST, a universal framework that integrates retrieval-augmented mechanisms with spatio-temporal modeling to address these challenges. Our framework consists of three key designs: 1) Decoupled Encoder and Query Generator to capture decoupled spatial and temporal features and construct a fusion query via residual fusion; 2) Spatio-temporal Retrieval Store and Retrievers to maintain and retrieve vectorized fine-grained patterns; and 3) Universal Backbone Predictor that flexibly accommodates pre-trained STGNNs or simple MLP predictors. Extensive experiments on 6 real-world traffic networks, including large-scale datasets, demonstrate that RAST achieves superior performance while maintaining computational efficiency.

Recent advancements in time series forecasting have explored augmenting models with text or vision modalities to improve accuracy. While text provides contextual understanding, it often lacks fine-grained temporal details. Conversely, vision captures intricate temporal patterns but lacks semantic context, limiting the complementary potential of these modalities. To address this, we propose Time-VLM, a novel multimodal framework that leverages pre-trained Vision-Language Models (VLMs) to bridge temporal, visual, and textual modalities for enhanced forecasting. Our framework comprises three key components: (1) a Retrieval-Augmented Learner, which extracts enriched temporal features through memory bank interactions; (2) a Vision-Augmented Learner, which encodes time series as informative images; and (3) a Text-Augmented Learner, which generates contextual textual descriptions. These components collaborate with frozen pre-trained VLMs to produce multimodal embeddings, which are then fused with temporal features for final prediction. Extensive experiments demonstrate that Time-VLM achieves superior performance, particularly in few-shot and zero-shot scenarios, thereby establishing a new direction for multimodal time series forecasting. Code is available at https://github.com/CityMind-Lab/ICML25-TimeVLM.

Climate change and rapid urbanization have led to the Urban Heat Island (UHI) effect, resulting in higher temperatures in metropolitan areas and negatively impacting urban communities. Accurate UHI forecasting is crucial for identifying high-risk periods and locations, especially in cities with vulnerable populations. Current methods are limited by data granularity and inadequate modeling of regional thermodynamics, which affects both accuracy and spatio-temporal granularity. In this paper, we propose DeepUHI, a data-driven context-aware framework for modeling local thermodynamics based on the heat equation, alongside the SeoulTemp dataset, the first multi-modal dataset for UHI effect predictions at the street level. Our framework utilizes a heat decomposition method to represent urban thermodynamics through thermodynamic cycles and thermal flows, effectively integrating urban environmental data. Extensive experiments show that our framework improves accuracy in UHI effect prediction and warning tasks, outperforming leading models. We have integrated DeepUHI into our SeoUHI platform to provide hourly street-level UHI forecasting for Seoul. The code, platform, and dataset are accessible at https://github.com/CityMind-Lab/DeepUHI.

Predicting spatio-temporal traffic flow presents significant challenges due to complex interactions between spatial and temporal factors. Existing approaches often address these dimensions in isolation, neglecting their critical interdependencies. In this paper, we introduce theSpatio-TemporalUnitizedModel (STUM), a unified framework designed to capture both spatial and temporal dependencies while addressing spatio-temporal heterogeneity through techniques such as distribution alignment and feature fusion. It also ensures both predictive accuracy and computational efficiency. Central to STUM is the Adaptive Spatio-temporal Unitized Cell (ASTUC), which utilizes low-rank matrices to seamlessly store, update, and interact with space, time, as well as their correlations. Our framework is also modular, allowing it to integrate with various spatio-temporal graph neural networks through components such as backbone models, feature extractors, residual fusion blocks, and the predictor to collectively enhance forecasting outcomes. Experimental results across multiple real-world datasets demonstrate that STUM consistently improves prediction performance with minimal computational cost. These findings are further supported by hyperparameter optimization, ablation studies, and result visualization. We provide our source code for reproducibility athttps://github.com/RWLinno/STUM