Towards integrated and fine-grained traffic forecasting: A Spatio-Temporal Heterogeneous Graph Transformer approach

Li, Guangyue; Zhao, Zilong; Guo, Xiaogang; Tang, Luliang; Zhang, Huazu; Wang, Jinghan

doi:10.1016/j.inffus.2023.102063

Cited by 12 publications

(4 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In summary, no model from previous works can efficiently handle all the mentioned challenges because most face difficulties in capturing spatial-temporal dependencies or long-term predictions. Therefore, researchers developed models that combine spatial and temporal neural networks with the attention technique to benefit from their advantages, 29,30,32,33,47,48 and they succeeded to a large extent. For example, Tang and Zeng 29 proposed a hybrid model called spatial-temporal correlation graph convolutional networks (STCGCN).…”

Section: Spatial-temporal Attention Networkmentioning

confidence: 99%

“…Although these models have achieved satisfactory results, they remain insufficient, especially with the appearance of recurrent neural networks (RNNs), 24,25 convolutional neural networks (CNNs), 20 graph convolutional networks (GCNs), 2,26 attention 7,27 and transformers, 28 which achieved great results, particularly when combined. For this reason, researchers proposed models that combine different techniques to produce the best results 29–34 …”

Section: Introductionmentioning

confidence: 99%

“…For this reason, researchers proposed models that combine different techniques to produce the best results. [29][30][31][32][33][34] Most of the above methods faced difficulties in handling long-term predictions. While certain algorithms, like CNNs and GCNs, have trouble extracting temporal correlations, others, like RNNs and their variations, have trouble extracting spatial dependencies.…”

mentioning

confidence: 99%

See 2 more Smart Citations

A spatial‐temporal graph gated transformer for traffic forecasting

Bouchemoukha,

Zennir,

Alioua

2024

Trans Emerging Tel Tech

View full text Add to dashboard Cite

Accurate traffic forecasting is more necessary than ever for transportation departments, especially given its significant role in traffic planning, management, and control. However, most existing methods struggle to address complex spatial correlations on road networks, nonlinear temporal dynamics, and difficult long‐term prediction. This article proposes a novel spatial temporal graph gated transformer (STGGT) to overcome these challenges. The suggested model differs from Google's transformer because it uses a hybrid architecture that integrates graph convolutional networks (GCNs), attention, and gated recurrent units (GRUs) instead of solely relying on attention. Specifically, STGGT uses GCNs to extract spatial dependencies, utilizes attention and GRUs to extract temporal dependencies, and handle long‐term prediction. Experiments indicate that STGGT outperforms the state‐of‐the‐art baseline models on two real‐world traffic datasets of 9%–40%. The proposed model offers a promising solution for accurate traffic forecasting, simultaneously addressing the challenges of complex spatial correlations, nonlinear temporal dynamics, and long‐term prediction.

show abstract

Section: Spatial-temporal Attention Networkmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

A spatial‐temporal graph gated transformer for traffic forecasting

Bouchemoukha,

Zennir,

Alioua

2024

Trans Emerging Tel Tech

View full text Add to dashboard Cite

show abstract

“…Emphasizing the societal implications [9], leveraging predictive analytics transcends academic pursuits; while the complexity of accident prediction has been elaborated upon, the real-world gains are significant. Harnessing data-driven insights through modern machine learning (ML) techniques [10], such as DL and AI, is not just a technological leap; it is a step towards safer roadways, economic efficiency, and, most importantly, the preservation of human life.…”

Section: Introduction and Related Workmentioning

confidence: 99%

LLM Multimodal Traffic Accident Forecasting

de Zarzà,

de Curtò,

Roig

et al. 2023

Sensors

View full text Add to dashboard Cite

With the rise in traffic congestion in urban centers, predicting accidents has become paramount for city planning and public safety. This work comprehensively studied the efficacy of modern deep learning (DL) methods in forecasting traffic accidents and enhancing Level-4 and Level-5 (L-4 and L-5) driving assistants with actionable visual and language cues. Using a rich dataset detailing accident occurrences, we juxtaposed the Transformer model against traditional time series models like ARIMA and the more recent Prophet model. Additionally, through detailed analysis, we delved deep into feature importance using principal component analysis (PCA) loadings, uncovering key factors contributing to accidents. We introduce the idea of using real-time interventions with large language models (LLMs) in autonomous driving with the use of lightweight compact LLMs like LLaMA-2 and Zephyr-7b-α. Our exploration extends to the realm of multimodality, through the use of Large Language-and-Vision Assistant (LLaVA)—a bridge between visual and linguistic cues by means of a Visual Language Model (VLM)—in conjunction with deep probabilistic reasoning, enhancing the real-time responsiveness of autonomous driving systems. In this study, we elucidate the advantages of employing large multimodal models within DL and deep probabilistic programming for enhancing the performance and usability of time series forecasting and feature weight importance, particularly in a self-driving scenario. This work paves the way for safer, smarter cities, underpinned by data-driven decision making.

show abstract