As urban data expands, existing spatio-temporal models encounter challenges such as high context dependency, poor cross-scenario generalization, and inefficient computational performance. To address these issues, we propose UrbanPG, an efficient and scalable framework for spatio-temporal learning. UrbanPG separates task-specific personalized patterns from general patterns, enabling unified spatio-temporal modeling and efficient knowledge generalization across scenarios. The key innovations of UrbanPG include: the development of a lightweight, context-independent general backbone utilizing linear spatio-temporal attention for scalable cross-scenario deployment; a personalized context prompt mechanism designed to model heterogeneity through spatio-temporal embeddings and random perturbation regularization, interacting with the backbone to enhance pattern differentiation; the proposal of multi spatio-temporal learning paradigms for rapid knowledge transfer and generalization to downstream tasks through fine-tuning personalized context prompts while freezing the backbone. Experimental results demonstrate that UrbanPG achieves state-of-the-art performance in large-scale forecasting, few-shot transfer, and continual learning tasks across eight real-world datasets, showcasing exceptional performance, strong generalization, and significant reductions in computational overhead.

Time series data plays a pivotal role in a wide variety of fields but faces challenges related to privacy concerns. Recently, synthesizing data via diffusion models is viewed as a promising solution. However, existing methods still struggle to capture long-range temporal dependencies and complex channel interrelations. In this research, we aim to utilize the sequence modeling capability of a State Space Model called Mamba to extend its applicability to time series data generation. We firstly analyze the core limitations in State Space Model, namely the lack of consideration for correlated temporal lag and channel permutation. Building upon the insight, we propose Lag Fusion Mamba and Permutation Scanning Mamba, which enhance the model's ability to discern significant patterns during the denoising process. Theoretical analysis reveals that both variants exhibit a unified matrix multiplication framework with the original Mamba, offering a deeper understanding of our method. Finally, we integrate two variants and introduce Diffusion Mamba for Time Series (DiM-TS), a high-quality time series generation model that better preserves the temporal periodicity and inter-channel correlations. Comprehensive experiments on public datasets demonstrate the superiority of DiM-TS in generating realistic time series while preserving diverse properties of data.

Next POI recommendation plays a crucial role in delivering personalized location-based services, but it faces significant challenges in capturing complex user behavior and adapting to dynamic interest distributions. Most methods often provide insufficient modeling of implicit features in user trajectories, such as directional transitions and latent edge relationships, which are essential for understanding user behavior. Moreover, existing diffusion models, constrained by Gaussian priors, struggle to handle the diverse and evolving nature of user preferences. The lack of a unified scheduling for noise and sampling also limits the flexibility of diffusion models. In this paper, we propose a Unified Bridge-based Diffusion model (UB-Diff) for the next POI recommendation. UB-Diff incorporates a direction-aware POI transition graph learning, which jointly captures spatio-temporal and directional features. To overcome the limitations of Gaussian priors, we introduce a bridge-based diffusion POI generative model. It can achieve distribution translation from the user's historical distribution to the target distribution by learning a bridge to associate user behavior with POI recommendation, adapting to dynamic user interests. In the end, we design a novel intermediate function to unify the diffusion process, enabling precise control over noise scheduling and modular optimization. Extensive experiments on five real-world datasets demonstrate the superiority of UB-Diff over advanced baseline methods. Our code is available athttps://github.com/JKZuo/UBDiff.

The contemporary urban intelligent transportation system (ITS) generates an enormous amount of trajectory data daily, serving as an essential reflection of traffic dynamics. Accurate estimation of arrival time by mining spatio-temporal features and semantic relationships from historical trajectories has become increasingly vital. However, most existing works overlook the joint features between links (i.e., road segments) and crossroads in trajectories. Additionally, they often treat all links uniformly without considering the semantics of critical links, leading to deficiencies in captured representation. To address these issues, this study proposes a novel deep encoder learning framework called the Triple Feature Encoder-based Dual-Granularity Graph Learning Network (TriDGNet) for enhanced travel time estimation. Specifically, we design a triple feature learning encoder to explore the spatio-temporal correlations of trajectories from three perspectives: Depth, Ensemble, and Sequence. Meanwhile, we introduce a consistent modeling method to integrate both links and crossroads. Furthermore, we construct two graph learning modules at different scales. One is an edge-enhanced graph attention network (E-GAT) to capture global spatial dependencies across the entire road network. The other is a backtracking-based subgraph representation network (BackNet) to learn local contextual information from bustling links. Our proposed TriDGNet model has been evaluated on three extensive datasets. The experimental results demonstrate that it outperforms state-of-the-art approaches.

Spatial-temporal forecasting is crucial and widely applicable in various domains such as traffic, energy, and climate. Benefiting from the abundance of unlabeled spatial-temporal data, self-supervised methods are increasingly adapted to learn spatial-temporal representations. However, it encounters three key challenges: 1) the difficulty in selecting reliable negative pairs due to the homogeneity of variables, hindering contrastive learning methods; 2) overlooking spatial correlations across variables over time; 3) limitations of efficiency and scalability in existing self-supervised learning methods. To tackle these, we propose a lightweight representation-learning model ST-ReP, integrating current value reconstruction and future value prediction into the pre-training framework for spatial-temporal forecasting. And we design a new spatial-temporal encoder to model fine-grained relationships. Moreover, multi-time scale analysis is incorporated into the self-supervised loss to enhance predictive capability. Experimental results across diverse domains demonstrate that the proposed model surpasses pre-training-based baselines, showcasing its ability to learn compact and semantically enriched representations while exhibiting superior scalability.

Urban spatio-temporal forecasting is critical for modern urban governance, especially in traffic management, resource planning, and emergency response. Despite advancements in pre-trained models for natural language processing, challenges persist in urban spatio-temporal forecasting. Existing methods struggle to identify and generalize universal cross-district spatio-temporal patterns, while computational limitations hinder the extraction of complex patterns from large-scale data. In this study, we propose CrossST, an efficient pre-training framework designed to capture universal spatio-temporal patterns across large-scale, cross-district scenarios. Specifically, CrossST performs pre-training on various large-scale spatio-temporal datasets to learn and store diverse valuable patterns in its pattern bank. It captures temporal dependencies, including periodicity and trends, through frequency domain and time domain analysis, while leveraging graph attention mechanisms to identify dynamic spatial propagation patterns. During fine-tuning, a spatio-temporal disentanglement strategy separates universal patterns from diverse spatio-temporal patterns stored during pre-training, improving generalization to downstream tasks and enabling efficient cross-district knowledge transfer. Additionally, temporal information aggregation and spatial linear optimization strategies enhance CrossST's efficiency and scalability, significantly reducing computational costs. Extensive experiments demonstrate that CrossST outperforms state-of-the-art baselines, improving downstream task generalization while maintaining low computational overhead. The datasets and code are available at https://github.com/Aoyu-Liu/CrossST.

Urban spatio-temporal prediction plays a crucial role in modern urban planning and management. However, spatio-temporal data exhibit complex and dynamic characteristics, presenting significant challenges for accurate forecasting. Although current state-of-the-art methods have made improvements, they still face two key limitations: (1) Many studies coarsely treat timestamps, failing to exploit the rich information embedded in timestamp data fully. (2) Most approaches overlook the differentiation of spatio-temporal patterns across urban nodes. Due to variations in the attributes of these nodes, even when local historical patterns are similar, future performance may differ. Thus, effective differentiation of spatio-temporal patterns is essential for accurate predictions. To address these challenges, we propose the STTP for urban spatio-temporal forecasting. Our approach introduces a trainable timestamp network to extract temporal correlation features from timestamps, using them as conditional prompts through an adaptive normalization technique to optimize spatio-temporal correlation capture. Additionally, we design a spatial attention module that differentiates node patterns, clustering similar patterns and differentiating dissimilar ones via cross-pattern and intra-pattern attention mechanisms. Extensive experimental results on six real-world datasets demonstrate that the STTP significantly outperforms existing state-of-the-art baselines. Codes are available at https://github.com/Dingct/STTP.

Accurate bus travel time prediction is essential for intelligent transportation systems, yet it remains challenging due to the nonlinear and dynamic nature of urban traffic. This paper proposes a novel hybrid model that combines a dual-stream LSTM encoder and a Transformer-based decoder within an encoder-decoder framework to capture both short- and long-term temporal dependencies. To effectively incorporate heterogeneous information, we design a Gated Feature Fusion Layer that adaptively integrates static and dynamic features. The decoder includes a Static-conditioned Variable Selection Block, Gated Residual Network, and Temporal Attention Layer to enhance feature extraction and temporal interpretability. Extensive experiments on a real-world dataset collected from Shanghai demonstrate that our model significantly outperforms baseline methods. Further ablation studies validate the effectiveness of each component, confirming the robustness and superiority of the proposed architecture in complex urban traffic scenarios.

Medical image generation plays an important role in clinical applications, particularly in modality translation tasks such as CT-to-MRI synthesis. Compared with supervised methods that rely on paired datasets, unsupervised image generation is more practical due to its lower data requirements. However, existing unsupervised methods often suffer from insufficient high-frequency detail preservation and poor style consistency. Although wavelet transform offers promising multi-scale representation capabilities, effectively integrating it into deep generative models remains challenging. To address these issues, we propose WDS-Net (Wavelet-based Dual-branch Style-consistent Network). In the content encoder, wavelet decomposition is used to separate the image into low-frequency structural and high-frequency detail components, which are processed through a dual-branch architecture and then fused to enhance content representation. A multi-scale feature extraction mechanism is employed in the style encoder, and layer-wise style injection is applied during decoding to improve style consistency. Experimental results demonstrate that WDS-Net can generate high-quality MRI images under unpaired training conditions and achieves robust performance even with limited data. Evaluations on both public and clinical datasets confirm that WDS-Net outperforms existing methods in detail preservation and style consistency, showing strong potential for real-world clinical applications.

This work explores the design of stick-shaped Tangible User Interfaces (TUI) for Extended Reality (XR). While sticks are widely used in everyday objects, their applications as a TUI in XR have not been systematically studied. We conducted a participatory design session with twelve experts in XR and human-computer interaction to investigate the affordances of stick-based objects and how to utilize them in XR. The results led us to develop a taxonomy of stick-based objects’ affordances in terms of their functions and holding gestures. Following that, we proposed four types of stick-based XR controller forms and discussed their advantages and limitations. In the end, we juxtaposed twenty-six existing XR controllers against our proposed forms and identified Landed (Cane) Stick, Thin Stick’s flexible usages, and Modular Design as the major opportunities that remain unexamined yet for stick-based XR TUI design.

Multivariate time series forecasting is crucial across various applications. In recent years, numerous studies adopt embedding layer and Attention mechanism to extract the intricate spatio-temporal features of time series. This involves directly transmitting the concatenated embeddings into the Attention mechanism. However, they generally overlook the importance of sending the integrated information in the embeddings into the Attention mechanism in a more appropriate way. To address this, we propose an intuitive network model with Temporal MLP Bridging the gap between Embedding and Attention (TMBEA) to deal with the above issue. Specifically, we explore a light-weight bridge with simple Multi-Layer Perceptrons (MLPs) fusing features along the temporal dimension, processing the embeddings before feeding them into the canonical Attention networks, which help embeddings to better align with the subsequent Attention networks. Experiments on real-world datasets, traffic datasets and air pollutant concentration datasets, demonstrate the efficiency of model. Further studies also show the capacity of bridge in improving the robustness of the model.

Effective Point-of-Interest (POI) recommendation systems play a pivotal role in modern location-aware applications and human mobility, facilitating customized suggestions for users’ upcoming exploration destinations. Understanding the intricate dynamics of user movement, which are often influenced by a multitude of factors, remains a formidable task. Moreover, discrepancies between the acquired representation distribution and the authentic target distribution of user interests also present a notable obstacle. To tackle these problems, we make an attempt to bridge the gap by introducing diffusion models and propose a diffusion-based dual graph multiattention network (Diff-DGMN). Specifically, we have constructed two types of graphs: one is a user-oriented local POI transition graph and the other is a global-based POI distance graph. Subsequently, we put forward two graph learning representation modules to capture the sequential encoding of users and the geographic representations of nodes, respectively. Furthermore, an attention-based location prototype generation module is introduced to merge the captured sequential encoding and geographic representation, yielding richer semantic interaction features. In the end, we obtain the final results by leveraging the forward diffusion process and corresponding its reverse-time generation to sample users’ future preferences from the posterior distribution. Our Diff-DGMN model demonstrates its remarkable recommendation performance through extensive experimentation on five real-world data sets. Compared with the most state-of-the-art methodologies, Diff-DGMN has improved performance in accuracy, normalized discounted cumulative gain (NDCG), and mean reciprocal rank (MRR) by 8.04%, 8.63%, and 9.09%, respectively. Our codes are available athttps://github.com/JKZuo/Diff-DGMN.

Accurate traffic forecasting is essential in urban traffic management, route planning, and flow detection. Recent advances in spatial-temporal models have markedly improved the modeling of intricate spatial-temporal correlations for traffic forecasting. Unfortunately, most previous studies have encountered challenges in effectively modeling spatial-temporal correlations across various perceptual perspectives and have neglected the interactive learning between spatial and temporal correlations. Additionally, constrained by spatial heterogeneity, most studies fail to consider distinct spatial-temporal patterns of each node. To overcome these limitations, we propose a Spatial-Temporal Interactive Dynamic Graph Convolutional Network (STIDGCN) for traffic forecasting. Specifically, we propose an interactive learning framework composed of spatial and temporal modules for downsampling traffic data. This framework aims to capture spatial and temporal correlations by adopting a perception perspective from the global to the local level and facilitating their mutual utilization with positive feedback. In the spatial module, we design a dynamic graph convolutional network based on graph construction methods. The network is designed to leverage a traffic pattern bank considering spatial-temporal heterogeneity as a query to reconstruct a data-driven dynamic graph structure. The reconstructed graph structure can reveal dynamic associations between nodes in the traffic network. Extensive experiments on eight real-world traffic datasets demonstrate that STIDGCN outperforms the state-of-the-art baseline while balancing computational costs. The source codes are available at https://github.com/LiuAoyu1998/STIDGCN.

Traffic forecasting plays a crucial role in intelligent transportation systems and finds application in various domains. Accurate traffic forecasting remains challenging due to the time-varying correlations within the data and the heterogeneous correlations between regions. Although various dynamic spatial-temporal graph models have been proposed to address these challenges in recent years, most of them are burdened by high computation costs and not intuitive to understand. In this paper, we propose a spatial-temporal graph model, Spatial-Temporal Dynamic Graph Diffusion Convolutional Network (SDGDN) that provides an effective and efficient approach to traffic forecasting. From the perspective of traffic flow transition probabilities, SDGDN learns dynamic graph structures to capture the time-varying traffic transition relationships. Besides dynamic graph structures, static node features are employed in diffusion convolution to better capture heterogeneous regional features. Furthermore, we utilize temporal encoding and also generate varying graphs in each stacked layer to enhance the forecasting performance. Experiments results on five real-world datasets demonstrate that SDGDN outperforms most baseline models in terms of both performance and computation efficiency.

Traffic flow forecasting is indispensable in modern urban life. Considering the complexity, variability and strong timeliness of traffic flow, traffic flow forecasting is a worth exploring but challenging research field. To achieve better traffic flow forecasting effect, we focus on two critical aspects that assume noteworthy importance: i) the features inside the traffic outflows and inflows. ii) the supplementary information regarding exterior region which is the area outside the grid division regions. To address these challenges, we propose a novel deep learning model Spatial-Temporal Flow Holistic Interaction Graph Convolution Network (STHGCN). In STHGCN, graph convolution based modules are applied through multi-step simulation. An exterior region feature estimation module is designed to estimate the influence of the special exterior region through the characteristics of complete trajectories, which enables a more comprehensive reasoning for traffic flow forecasting in grid division regions. Furthermore, a flow feature fusion integrator and stackable convolution modules are proposed to aggregate the intermediate features extracted from various perspectives, which simulate the constantly-updating and interlinked states of traffic flows through the process of multi-layer feature separation and fusion. We conduct extensive experiments on real-world traffic datasets and our proposed model outperforms all baselines.

In recent years, the travel time prediction has been receiving sustained attention because of the prevalence of location based applications, smart city engineering and online car-hailing. In this paper, a Self-organizing Graph Embedding Deep Network (SGED-Net) model is proposed to address the challenging Origin-Destination(OD) based travel time estimation (TTE). Specifically, SGED-Net comprises four modules, involving travel feature extraction, spatial association graph generation, graph embedding, and travel time deep learning prediction modules. First, we comprehensively extract three travel characteristics and feed them into a modified LightGBM component for learning the importance sort of features for different trips. Second, considering the lack of intermediate trajectories in the OD-based TTE, we employ a self-organizing feature mapping (SOM) approach to obtain the prominent nodes among pick-up and drop-off locations. Meanwhile clustering algorithm is used to divide a city into n-clusters districts of variant sizes. The topology learned by SOM is combined with the districts divided by clustering algorithm to obtain a district-based spatial association graph. Moreover, an improved SDNE algorithm is leveraged to gain a low-dimension spatial association representation while preserving the global and local structure of the graph. Then, we design a deep neural network for learning the captured spatial association representation. Finally, a series of experiments in two real-world large-scale datasets demonstrate the SGED-Net's excellent performance.

Capturing complex and dynamic spatial-temporal dependencies of traffic data is of great importance for accurate and real-time traffic forecasting in intelligent transportation systems. The spatial-temporal dependency between traffic locations is often dynamic, which means the correlation between the traffic status of different locations changes jointly over their spatial distance and the time slice they are in. Most of the existing Graph Convolutional Network-based methods usually capture spatial and temporal dependencies separately and then combine them in a parallel or serial mechanism to capture the spatial-temporal features. They always utilize the predefined static graph structure to capture both local correlations and global dependencies in the same time slice. These methods are incapable of directly learning dynamic spatial-temporal dependency across time slices. Meanwhile, it is challenging to learn the spatiotemporal correlation knowledge among traffic locations only by using neural networks. To address these issues, we propose a novel Dynamic Spatial-Temporal Adjacent Graph Convolutional Network (DSTAGCN), which connects the latest time slice with each past time slice to construct the spatial-temporal graph. DSTAGCN can directly learn the global spatial dependency across time and simultaneously capture the spatial-temporal dependencies through graph convolution. Since fuzzy theory make it possible to represent uncertain relationships, a simplified fuzzy neural network that integrates fuzzy systems and neural networks is designed to help generating the graph adjacency matrix representing the dynamic adjacency correlations. Experiments on public datasets show that our method outperforms baselines with fast convergence.

With the enhancement of location-acquisition technologies, GPS trajectories play an essential role in data-driven intelligent transportation applications, which requires an accurate approach to match raw GPS trajectories to road segments on a digital map. However, for complex urban roads containing elevated roads and surface roads, map matching for low-frequency GPS data is still challenging. This article aims to address the biases and instability problem in existing approaches. To this end, we combine the spatial-temporal characteristics of GPS data in complex roads with driving behaviours and present a novel global map matching method including truncated density clustering algorithm, statistic features based spatial-temporal analysis, and driving-behaviour-based track modification. Additionally, a weighted-matrix based interactive voting algorithm is proposed to select the best results from a global perspective. The experiments are conducted with two real GPS trajectory datasets under three road conditions. The result shows that our approach outperforms state-of-art approaches for urban complex road networks in both accuracy and efficiency.

Path recommendation is an essential application in people’s daily life. However, drivers’ experience hidden in their driving history and their personal preferences are less considered in path planning. In this paper we propose a dynamic time-constrained path recommendation method, utilizing the historical GPS trajectory information and user preferences. First, based on graph entropy, critical intersections are extracted from trajectory information to simplify the road network structure. Then the time-dependent mostly chosen path (TDMCP) is obtained by trajectories processing to determine the time-varying paths between critical intersections. Thus the original complex road network is abstracted as a search subnet. Finally we propose an improved dynamic A* search (IDAS) on the search sub-net to find the candidate paths satisfying the time-constraints, and further allow users to set personal preferences to select the final optimal path. The proposed method is validated on the real-world data set, and the result shows that the proposed method surpasses the competing ones especially on long distance path recommendation.

The world is entering a new normal of hybrid organisations, in which it will be common that some members are co-located and others are remote. Hybridity is rife with asymmetries that affect our sense of belonging in an organisational space. This paper reports a study of an XR Telepresence technology probe to explore how remote workers might present themselves and be perceived as an equal and unique embodied being in a workplace. VROOM (Virtual Robot Overlay for Online Meetings) augments a standard Mobile Robotic Telepresence experience by (1) adding a virtual avatar overlay of the remote person to the local space, viewable through a HoloLens worn by the local user, through which the remote user can gesture and express themselves, and (2) giving the remote user an immersive 360° view of the local space, captured by a 360° camera on the robot, which they can view through a VR headset. We ran a study to understand how pairs of participants (one local and one remote) collaborate using VROOM in a search and word-guessing game. Our findings illustrate that there is much potential for a system like VROOM to support dynamic collaborative activities in which embodiment, gesturing, mobility, spatial awareness, and non-verbal expressions are important. However, there are also challenges to be addressed, specifically around proprioception, the mixing of a physical robot body with a virtual human avatar, uncertainties of others' views and capabilities, fidelity of expressions, and the appearance of the avatar. We conclude with further design suggestions and recommendations for future work.

Many studies have proposed different ways of supporting flying in embodied virtual reality (VR) interfaces with limited success. Our research explores the usage of a user's lower body to support flying locomotion control through a novel "flexible perching" (FlexPerch) stance that provides user with leg moving ability while sitting. We conducted an observational study exploring participants' preferred usage of the FlexPerch stance, and a mixed-method study comparing the same flying experience with existing sitting and standing stances. Our results show that FlexPerch markedly increased participants' feelings of flying. However, people may not like "flying" when they really can - the freedom, feeling of floating, and novelty contributing to this sensation can also mean more effort and feeling unsafe or unfamiliar. We suggest that researchers studying VR flying interfaces evaluate the feeling of flying, and raise design considerations to use stances like FlexPerch to elicit feelings of flying and stimulation.

The urban expressway system model based on hybrid petri nets is proposed. With the traffic detecting cameras and coils, the strategies of traffic signal lights and warning lights are enforced with the aim to prevent the large-scale congestion in emergencies (e.g. accidents). The simulation results verify the effectiveness of the proposed model and the warning light strategy is demonstrated to be suitable for traffic accidents and congestion in the accident-prone weaving sections in the urban expressways.

A vehicle tracking mode based on dynamic roles is proposed in the smart camera networks. The tracking for a specific vehicle is organized collaboratively and automatically in smart camera network. The tracking is completed autonomously by smart cameras in a distributed manner. The decentralized control mode can effectively decrease the working pressure of server system, reduce the requirements of network bandwidth for real-time video transmission, and make the system flexible and fault-tolerant. An information transmission mechanism is also presented in the smart camera network to ensure the collaborative tracking. Finally, a demo system for vehicle tracking in intelligent traffic surveillance system is implemented to verify the proposed methods.

Improving driving safety is one major objective of forming vehicular ad hoc networks (VANETs). Existing VANETs usually assume drivers detect and report safety-related road conditions. However, drivers may not be willing to perform these duties; even they are, these duties may distract them from driving and thus make driving unsafe. To address the problem, this paper proposes an automatic detection system. By taking advantage of the communication capability of roadside sensors, the proposed system can automatically detect and locate problematic road conditions without any human intervention under varying traffic densities. Extensive simulations have been conducted to verify the efficiency of the proposed system.

The current stepwise refinement operation of Petri nets mainly concentrates on property preservation, which is an effective way to analyze and verify complex systems. Further steps into this field are needed from the perspective of system synthesis and language preservation. First, the refinement of Petri nets is introduced based on a$k$-well-behaved Petri net, in which$k$tokens can be processed. Then, according to the different compositions of subsystems, well-, under- and overmatched refined Petri nets are proposed. In addition, the language and property relationships among sub-, original, and refined nets are studied to demonstrate behavior characteristics and property preservation in a system synthesis process. A manufacturing system is given as an example to illustrate the effectiveness of the proposed approach in synthesizing and analyzing the Petri nets of complex systems.