A Bayesian generalised extreme value model to estimate real-time pedestrian crash risks at signalised intersections using artificial intelligence-based video analytics

Ali, Yasir; Haque, Md. Mazharul; Mannering, Fred

doi:10.1016/j.amar.2022.100264

Cited by 26 publications

(8 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Further, this comparison also reveals the relatively superior performance of Model-2, implying that a model with covariates in the location parameter outperforms its competing models. This finding implies that the model with covariates in the location parameter provides a theoretical underpinning by informing how close or far the zero modified time-to-collision point is on the curve relative to the mean 6 . Similar findings were obtained for the other two study intersections.…”

Section: Resultsmentioning

confidence: 87%

“…These three datasets are processed simultaneously to obtain signal cycle level covariates. For this purpose, this study adapts the automated covariate extraction algorithm 6 for rear-end conflicts. At the core of this algorithm is a data fusion technique combining three datasets to extract covariates at the signal cycle level.…”

Section: Data and Pre-processingmentioning

confidence: 99%

“…Whilst EVT assists in establishing the link between conflict extremes and crashes at the micro level, like a signal cycle, the time-series forecasting capability of recurrent neural networks has been leveraged to forecast the crash risk based on past and present crash risk. An automated covariate extraction algorithm for the signal cycle, proposed by Ali et al 6 , is extended to calculate the covariates for rear-end conflicts automatically. This algorithm uses data fusion principles for combining three data sources.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics

Hussain,

Ali,

et al. 2024

Sci Rep

View full text Add to dashboard Cite

This study proposes a bi-level framework for real-time crash risk forecasting (RTCF) for signalised intersections, leveraging the temporal dependency among crash risks of contiguous time slices. At the first level of RTCF, a non-stationary generalised extreme value (GEV) model is developed to estimate the rear-end crash risk in real time (i.e., at a signal cycle level). Artificial intelligence techniques, like YOLO and DeepSort were used to extract traffic conflicts and time-varying covariates from traffic movement videos at three signalised intersections in Queensland, Australia. The estimated crash frequency from the non-stationary GEV model is compared against the historical crashes for the study locations (serving as ground truth), and the results indicate a close match between the estimated and observed crashes. Notably, the estimated mean crashes lie within the confidence intervals of observed crashes, further demonstrating the accuracy of the extreme value model. At the second level of RTCF, the estimated signal cycle crash risk is fed to a recurrent neural network to predict the crash risk of the subsequent signal cycles. Results reveal that the model can reasonably estimate crash risk for the next 20–25 min. The RTCF framework provides new pathways for proactive safety management at signalised intersections.

show abstract

Section: Resultsmentioning

confidence: 87%

Section: Data and Pre-processingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics

Hussain,

Ali,

et al. 2024

Sci Rep

View full text Add to dashboard Cite

show abstract

“…Second, crashes estimated using EVT rely on traffic conflict and fit well within Hydén's pyramid (Hydén, 1987), suggesting that traffic extremes identified through conflicts can provide insights into the crash mechanism. Several studies confirm the suitability of EVT models for crash predictions (e.g., Ali et al (2023), Ali et al (2021a), Ali et al (2021b), Arun et al (2021b), Arun et al (2021c), Mohammadian et al (2021), and Zheng and Sayed (2020)). It is worth noting that Zheng and Sayed (2020) analyzed crash risk at a signal cycle; however, their study did not shed light on how crash risk varies across periods and whether these differences are statistically significant.…”

Section: Introductionmentioning

confidence: 71%

Estimating Real-Time Crash Risk at Signalized Intersections: A Bayesian Generalized Extreme Value Approach

2022

View full text Add to dashboard Cite

“…Therefore, more research is needed on the manner in which various SSMs could be exploited to enhance the safety of VRUs. Towards this direction, Ali et al developed a Bayesian Generalized EVT model in order to estimate real-time pedestrian crash risks at signalized intersections using Artificial Intelligence (AI)-based video analytics [92].…”

Section: Future Research Directionsmentioning

confidence: 99%

A Review of Surrogate Safety Measures Uses in Historical Crash Investigations

2023

View full text Add to dashboard Cite

Historical road crash data are the main indicator for measuring road safety outcomes. Over the past few decades, significant efforts have been made in obtaining and exploiting Surrogate Safety Measures (SSMs). SSMs have the potential to provide excellent sustainable road safety indicators and proxy measurements which can complement traditional historical crash analyses or even substitute them. By using SSMs, crash data collection demands can be bypassed and areas can be investigated before crashes occur. Due to such advantages, the objective of the present research is to provide a review of the scientific literature regarding studies exploiting SSMs for historical crash record investigations. Specifically, 34 studies were examined, providing insights on the different types of SSMs collected under real road environment conditions, the way they are collected, their connection with specific road crash types, and the type of the developed statistical models are examined and discussed. Particular focus is also placed on the temporal dimension of the collection period of both SSMs and road crashes. Finally, the overall trends deriving from the reviewed studies are summarized and future research directions are provided.

show abstract

A Bayesian generalised extreme value model to estimate real-time pedestrian crash risks at signalised intersections using artificial intelligence-based video analytics

Cited by 26 publications

References 47 publications

A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics

A bi-level framework for real-time crash risk forecasting using artificial intelligence-based video analytics

Estimating Real-Time Crash Risk at Signalized Intersections: A Bayesian Generalized Extreme Value Approach

A Review of Surrogate Safety Measures Uses in Historical Crash Investigations

Contact Info

Product

Resources

About