Jinli Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jinli Liu has a total of 37 co-authors. Here are the top co-authors by number of shared publications:
- Subasish Das
Ingram School of Engineering, Texas State University, San Marcos, USA
19 shared publications - Rohit Chakraborty
Ingram School of Engineering, Texas State University
6 shared publications - Gian Antariksa
Texas State University
3 shared publications - Nasim Khan
Department of Civil & Architectural Engineering & Construction Management, University of Wyoming, Laramie, WY
3 shared publications - Bahar Dadashova
Texas A&M Transportation Institute, College Station, TX 77843, USA
2 shared publications - Boniphace Kutela
Texas A&M Transportation Institute, Houston, TX
2 shared publications
Latest Publications
- 25 June 2026
- Published by Taylor & Francis in Journal of Transportation Safety & Security
- p. 1-33
Crash severity in automation-engaged vehicles: A comparative analysis across SAE levels
1. Road traffic crashes remain a pressing public health concern worldwide, claiming approximately 1.19 million lives each year and causing 20 to 50 million non-fatal injuries according to the World...
- 25 June 2026
- Published by Taylor & Francis in Journal of Transportation Safety & Security
- p. 1-33
Crash severity in automation-engaged vehicles: A comparative analysis across SAE levels
1. Road traffic crashes remain a pressing public health concern worldwide, claiming approximately 1.19 million lives each year and causing 20 to 50 million non-fatal injuries according to the World...
- Open Access
- 23 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 33665-33683
Image-Based Crash Severity Analysis: Transforming Tabular Data Using Convolutional Neural Networks
Traffic safety is a critical issue in urban transportation systems, with crash severity predictions playing a key role in mitigating road traffic injuries and fatalities. This study addresses the research question: how can transforming tabular crash data into image-based representations using deep learning techniques improve the prediction of crash severity, while effectively managing class imbalances? For analysis, 63,745 barrier crashes that occurred in Texas from 2017–2022 were examined. Using a sample database of barrier crash incidents, we introduce a novel approach that leverages Convolutional Neural Networks (CNNs) to convert traditional tabular crash data into image representations by employing the DeepInsight technique, which maps feature similarity into structured two-dimensional grids, enhancing the accuracy and effectiveness of prediction models. Generative Adversarial Networks (GANs) were used for data augmentation, addressing issues of class imbalances with innovative sampling techniques including under-sampling, over-sampling, and balanced sampling. Key findings indicate that balanced datasets significantly improve predictive performance, enhancing generalizability across diverse crash scenarios. Moreover, under- and over-sampling strategies offer a refined trade-off among precision, recall, and F1-score, while the over-sampling method distinctly excels in practical model performance by reducing misclassification rates and improving overall accuracy. Comparative evaluation against an XGBoost baseline on identical balanced data highlights the ability of the image-based deep learning framework to capture complex crash patterns. These findings suggest that integrating feature-similarity–based tabular-to-image transformation with deep learning and balanced data sampling can significantly improve crash severity classification.
- 23 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 33665-33683
Open AccessImage-Based Crash Severity Analysis: Transforming Tabular Data Using Convolutional Neural Networks
Traffic safety is a critical issue in urban transportation systems, with crash severity predictions playing a key role in mitigating road traffic injuries and fatalities. This study addresses the research question: how can transforming tabular crash data into image-based representations using deep learning techniques improve the prediction of crash severity, while effectively managing class imbalances? For analysis, 63,745 barrier crashes that occurred in Texas from 2017–2022 were examined. Using a sample database of barrier crash incidents, we introduce a novel approach that leverages Convolutional Neural Networks (CNNs) to convert traditional tabular crash data into image representations by employing the DeepInsight technique, which maps feature similarity into structured two-dimensional grids, enhancing the accuracy and effectiveness of prediction models. Generative Adversarial Networks (GANs) were used for data augmentation, addressing issues of class imbalances with innovative sampling techniques including under-sampling, over-sampling, and balanced sampling. Key findings indicate that balanced datasets significantly improve predictive performance, enhancing generalizability across diverse crash scenarios. Moreover, under- and over-sampling strategies offer a refined trade-off among precision, recall, and F1-score, while the over-sampling method distinctly excels in practical model performance by reducing misclassification rates and improving overall accuracy. Comparative evaluation against an XGBoost baseline on identical balanced data highlights the ability of the image-based deep learning framework to capture complex crash patterns. These findings suggest that integrating feature-similarity–based tabular-to-image transformation with deep learning and balanced data sampling can significantly improve crash severity classification.
- 1 August 2025
- Published by Taylor & Francis in Journal of Transportation Safety & Security
- Vol. 17 (12) , 1377-1417
Impact of lighting conditions on nighttime crash severity among older and elderly drivers
- 1 August 2025
- Published by Taylor & Francis in Journal of Transportation Safety & Security
- Vol. 17 (12) , 1377-1417
Impact of lighting conditions on nighttime crash severity among older and elderly drivers
- 9 July 2025
- Published by Taylor & Francis in Journal of Transportation Safety & Security
- Vol. 17 (11) , 1290-1318
Clustering patterns of roadway departure related motorcycle crashes using dimension reduction analysis
- 9 July 2025
- Published by Taylor & Francis in Journal of Transportation Safety & Security
- Vol. 17 (11) , 1290-1318
Clustering patterns of roadway departure related motorcycle crashes using dimension reduction analysis
- 26 June 2025
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2679 (9) , 967-980
COVID-19 Era Crash Fatality/Severe Injury and Proven Speed–Crash Relations
The number of highway crash fatalities increased significantly in 2020 compared to the previous year, coinciding with the unique traffic conditions brought about by the COVID-19 pandemic. This research investigates the hypothesis that increased speeds, influenced by various factors during the pandemic, may have contributed to the elevated fatality rates. Speed is widely recognized as a critical factor affecting crash severity. To test this hypothesis, the research team analyzed speed changes and their relationships with fatal and serious injury crashes using Washington State data specifically from 2018 to 2020. The results showed that despite a reduction in total crashes and vehicle miles traveled in 2020, fatal and serious injury crash rates per million vehicle miles traveled increased compared to 2017 and 2018, indicating a shift toward higher-severity crashes during the COVID-19 pandemic. Average operating speeds increased across all facility types in 2020, with a corresponding decrease in speed variability, particularly on urban roadways, suggesting smoother but faster traffic flow during the pandemic. The findings confirm that increased operating speeds during the pandemic were associated with higher fatal and serious injury crash rates. The conclusion of this research could help highway safety planners and designers better understand relationships between speed increase and safety during the pandemic and, by planning to manage speed, could decrease fatalities resulting from highway crashes.
- 26 June 2025
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2679 (9) , 967-980
COVID-19 Era Crash Fatality/Severe Injury and Proven Speed–Crash Relations
The number of highway crash fatalities increased significantly in 2020 compared to the previous year, coinciding with the unique traffic conditions brought about by the COVID-19 pandemic. This research investigates the hypothesis that increased speeds, influenced by various factors during the pandemic, may have contributed to the elevated fatality rates. Speed is widely recognized as a critical factor affecting crash severity. To test this hypothesis, the research team analyzed speed changes and their relationships with fatal and serious injury crashes using Washington State data specifically from 2018 to 2020. The results showed that despite a reduction in total crashes and vehicle miles traveled in 2020, fatal and serious injury crash rates per million vehicle miles traveled increased compared to 2017 and 2018, indicating a shift toward higher-severity crashes during the COVID-19 pandemic. Average operating speeds increased across all facility types in 2020, with a corresponding decrease in speed variability, particularly on urban roadways, suggesting smoother but faster traffic flow during the pandemic. The findings confirm that increased operating speeds during the pandemic were associated with higher fatal and serious injury crash rates. The conclusion of this research could help highway safety planners and designers better understand relationships between speed increase and safety during the pandemic and, by planning to manage speed, could decrease fatalities resulting from highway crashes.
- 18 April 2025
- Published by Elsevier in Accident Analysis & Prevention
- Vol. 217, 108048
Evaluating crash risk factors of farm equipment vehicles on county and non-county roads using interpretable tabular deep learning (TabNet)
- 18 April 2025
- Published by Elsevier in Accident Analysis & Prevention
- Vol. 217, 108048
Evaluating crash risk factors of farm equipment vehicles on county and non-county roads using interpretable tabular deep learning (TabNet)
- 20 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 26 (4) , 5568-5584
Automated and Explainable Artificial Intelligence to Enhance Prediction of Pedestrian Injury Severity
This study used a detailed explainable AI for automatic machine learning, AutoGluon, to predict pedestrian injury severity using data collected over five years (2016-2021) in Louisiana. The final dataset includes forty variables related to pedestrian characteristics, environmental circumstances, and vehicle specifications. Pedestrian injury severity was divided into three categories: fatal, injury, and no injury. The novelty of this approach lies in the application of explainable AI (XAI), specifically SHAP (SHapley Additive exPlanations) values, to interpret the AutoML model’s predictions. This combination not only addressed the opaqueness typically associated with AI “black box” models but also illuminated the critical variables influencing pedestrian injury severity outcomes. The results revealed that the weighted ensemble model emerged as top performers, showcasing high accuracy with minimal prediction times, demonstrating the potential of ensemble methods in improving prediction outcomes by integrating the strengths of various individual models. Furthermore, the global and local explainability analyses provided by SHAP values afforded us an in-depth understanding of the variables influencing pedestrian injury severity. This dual-level explanation offered valuable insights into the complex dynamics at play, ranging from pedestrian impairment and driver condition to environmental variables like lighting and weather conditions. These findings underscore the importance of specific variables in crash outcomes, offering actionable intelligence for targeted interventions.
- 20 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 26 (4) , 5568-5584
Automated and Explainable Artificial Intelligence to Enhance Prediction of Pedestrian Injury Severity
This study used a detailed explainable AI for automatic machine learning, AutoGluon, to predict pedestrian injury severity using data collected over five years (2016-2021) in Louisiana. The final dataset includes forty variables related to pedestrian characteristics, environmental circumstances, and vehicle specifications. Pedestrian injury severity was divided into three categories: fatal, injury, and no injury. The novelty of this approach lies in the application of explainable AI (XAI), specifically SHAP (SHapley Additive exPlanations) values, to interpret the AutoML model’s predictions. This combination not only addressed the opaqueness typically associated with AI “black box” models but also illuminated the critical variables influencing pedestrian injury severity outcomes. The results revealed that the weighted ensemble model emerged as top performers, showcasing high accuracy with minimal prediction times, demonstrating the potential of ensemble methods in improving prediction outcomes by integrating the strengths of various individual models. Furthermore, the global and local explainability analyses provided by SHAP values afforded us an in-depth understanding of the variables influencing pedestrian injury severity. This dual-level explanation offered valuable insights into the complex dynamics at play, ranging from pedestrian impairment and driver condition to environmental variables like lighting and weather conditions. These findings underscore the importance of specific variables in crash outcomes, offering actionable intelligence for targeted interventions.