Towards Effective Fusion and Forecasting of Multimodal Spatio-temporal Data for Smart Mobility
With the rapid development of location based services, multimodal spatio-temporal (ST) data including trajectories, transportation modes, traffic flow and social check-ins are being collected for deep learning based methods. These deep learning based methods learn ST correlations to support the downstream tasks in the fields such as smart mobility, smart city and other intelligent transportation systems. Despite their effectiveness, ST data fusion and forecasting methods face practical challenges in real-world scenarios. First, forecasting performance for ST data-insufficient area is inferior, making it necessary to transfer meta knowledge from heterogeneous area to enhance the sparse representations. Second, it is nontrivial to accurately forecast in multi-transportation-mode scenarios due to the fine-grained ST features of similar transportation modes, making it necessary to distinguish and measure the ST correlations to alleviate the influence caused by entangled ST features. At last, partial data modalities (e.g., transportation mode) are lost due to privacy or technical issues in certain scenarios, making it necessary to effectively fuse the multimodal sparse ST features and enrich the ST representations. To tackle these challenges, our research work aim to develop effective fusion and forecasting methods for multimodal ST data in smart mobility scenario. In this paper, we will introduce our recent works that investigates the challenges in terms of various real-world applications and establish the open challenges in this field for future work.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Multimodal Spatiotemporal-Frequency Fusion with Peak Enhancement for Cellular Traffic Forecasting
Accurate forecasting of cellular network traffic is essential for network planning, resource allocation, and quality-of-service assurance in modern mobile communication systems. Real-world traffic often exhibits bursty e…
CRISP: A Spatiotemporal Camera-Radar Backbone for Driving via Forecasting-Based World-Model Pretraining
Camera-radar (CR) fusion is a practical sensing configuration for autonomous driving, but existing models are typically trained with task-specific supervision, limiting reusable representation learning. We present CRISP,…
Representation LearningAutonomous DrivingMotion ForecastingPoint CloudsGenerative diffusion models for spatiotemporal influenza forecasting
Forecasting infectious disease incidence can provide important information to guide public health planning, yet is difficult because epidemic dynamics are complex. Current mechanistic and statistical approaches often str…
ProGen: Revisiting Probabilistic Spatial-Temporal Time Series Forecasting from a Continuous Generative Perspective Using Stochastic Differential Equations
Accurate forecasting of spatiotemporal data remains challenging due to complex spatial dependencies and temporal dynamics. The inherent uncertainty and variability in such data often render deterministic models insuffici…
DenoisingTime SeriesTime Series ForecastingMulti-Spatio-temporal Fusion Graph Recurrent Network for Traffic forecasting
Traffic forecasting is essential for the traffic construction of smart cities in the new era. However, traffic data's complex spatial and temporal dependencies make traffic forecasting extremely challenging. Most existin…
ManagementTime Series Analysis