Hui Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Hui Chen has a total of 24 co-authors. Here are the top co-authors by number of shared publications:
Liming Lou
Research and Development Headquarters, JTEKT Co., Ltd, Japan
2 shared publications- Taokai Xia
School of Automotive Studies, Tongji University, Shanghai 201804, China
2 shared publications - Wei Ran
School of Automotive Studies, Tongji University, Shanghai 201804, China
2 shared publications - Bin Zhong
School of Automotive Studies, Tongji University, Shanghai 201804, China
1 shared publication Bolin Gao
Tongji University
1 shared publication- Chaopeng Guo
JTEKT Corporation, Nara 634-8555, Japan
1 shared publication
Latest Publications
- Open Access
A Survey of Autonomous Vehicle Behaviors: Trajectory Planning Algorithms, Sensed Collision Risks, and User Expectations
Autonomous vehicles are rapidly advancing and have the potential to revolutionize transportation in the future. This paper primarily focuses on vehicle motion trajectory planning algorithms, examining the methods for estimating collision risks based on sensed environmental information and approaches for achieving user-aligned trajectory planning results. It investigates the different categories of planning algorithms within the scope of local trajectory planning applications for autonomous driving, discussing and differentiating their properties in detail through a review of the recent studies. The risk estimation methods are classified and introduced based on their descriptions of the sensed collision risks in traffic environments and their integration with trajectory planning algorithms. Additionally, various user experience-oriented methods, which utilize human data to enhance the trajectory planning performance and generate human-like trajectories, are explored. The paper provides comparative analyses of these algorithms and methods from different perspectives, revealing the interconnections between these topics. The current challenges and future prospects of the trajectory planning tasks in autonomous vehicles are also discussed.
Open AccessA Survey of Autonomous Vehicle Behaviors: Trajectory Planning Algorithms, Sensed Collision Risks, and User Expectations
Autonomous vehicles are rapidly advancing and have the potential to revolutionize transportation in the future. This paper primarily focuses on vehicle motion trajectory planning algorithms, examining the methods for estimating collision risks based on sensed environmental information and approaches for achieving user-aligned trajectory planning results. It investigates the different categories of planning algorithms within the scope of local trajectory planning applications for autonomous driving, discussing and differentiating their properties in detail through a review of the recent studies. The risk estimation methods are classified and introduced based on their descriptions of the sensed collision risks in traffic environments and their integration with trajectory planning algorithms. Additionally, various user experience-oriented methods, which utilize human data to enhance the trajectory planning performance and generate human-like trajectories, are explored. The paper provides comparative analyses of these algorithms and methods from different perspectives, revealing the interconnections between these topics. The current challenges and future prospects of the trajectory planning tasks in autonomous vehicles are also discussed.
- Open Access
A Method to Develop the Driver-Adaptive Lane-Keeping Assistance System Based on Real Driver Preferences
To satisfy the preference of each driver, the development of a Lane-Keeping Assistance (LKA) system that can adapt to individual drivers has become a research hotspot in recent years. However, existing studies have mostly relied on the assumption that the LKA characteristic aligned with the driver’s preference is consistent with this driver’s naturalistic driving characteristic. Nevertheless, this assumption may not always hold true, causing limitations to the effectiveness of this method. This paper proposes a novel method for a Driver-Adaptive Lane-Keeping Assistance (DALKA) system based on drivers’ real preferences. First, metrics are extracted from collected naturalistic driving data using action point theory to describe drivers’ naturalistic driving characteristics. Then, the subjective and objective evaluation method is introduced to obtain the real preference of each test driver for the LKA system. Finally, machine learning methods are employed to train a model that relates naturalistic driving characteristics to the drivers’ real preferences, and the model-predicted preferences are integrated into the DALKA system. The developed DALKA system is then subjectively evaluated by the drivers. The results show that our DALKA system, developed using this method, can enhance or maintain the subjective evaluations of the LKA system for most drivers.
Open AccessA Method to Develop the Driver-Adaptive Lane-Keeping Assistance System Based on Real Driver Preferences
To satisfy the preference of each driver, the development of a Lane-Keeping Assistance (LKA) system that can adapt to individual drivers has become a research hotspot in recent years. However, existing studies have mostly relied on the assumption that the LKA characteristic aligned with the driver’s preference is consistent with this driver’s naturalistic driving characteristic. Nevertheless, this assumption may not always hold true, causing limitations to the effectiveness of this method. This paper proposes a novel method for a Driver-Adaptive Lane-Keeping Assistance (DALKA) system based on drivers’ real preferences. First, metrics are extracted from collected naturalistic driving data using action point theory to describe drivers’ naturalistic driving characteristics. Then, the subjective and objective evaluation method is introduced to obtain the real preference of each test driver for the LKA system. Finally, machine learning methods are employed to train a model that relates naturalistic driving characteristics to the drivers’ real preferences, and the model-predicted preferences are integrated into the DALKA system. The developed DALKA system is then subjectively evaluated by the drivers. The results show that our DALKA system, developed using this method, can enhance or maintain the subjective evaluations of the LKA system for most drivers.
- Open Access
Online Personalized Preference Learning Method Based on In-Formative Query for Lane Centering Control Trajectory
The personalization of autonomous vehicles or advanced driver assistance systems has been a widely researched topic, with many proposals aiming to achieve human-like or driver-imitating methods. However, these approaches rely on an implicit assumption that all drivers prefer the vehicle to drive like themselves, which may not hold true for all drivers. To address this issue, this study proposes an online personalized preference learning method (OPPLM) that utilizes a pairwise comparison group preference query and the Bayesian approach. The proposed OPPLM adopts a two-layer hierarchical structure model based on utility theory to represent driver preferences on the trajectory. To improve the accuracy of learning, the uncertainty of driver query answers is modeled. In addition, informative query and greedy query selection methods are used to improve learning speed. To determine when the driver’s preferred trajectory has been found, a convergence criterion is proposed. To evaluate the effectiveness of the OPPLM, a user study is conducted to learn the driver’s preferred trajectory in the curve of the lane centering control (LCC) system. The results show that the OPPLM can converge quickly, requiring only about 11 queries on average. Moreover, it accurately learned the driver’s favorite trajectory, and the estimated utility of the driver preference model is highly consistent with the subject evaluation score.
Open AccessOnline Personalized Preference Learning Method Based on In-Formative Query for Lane Centering Control Trajectory
The personalization of autonomous vehicles or advanced driver assistance systems has been a widely researched topic, with many proposals aiming to achieve human-like or driver-imitating methods. However, these approaches rely on an implicit assumption that all drivers prefer the vehicle to drive like themselves, which may not hold true for all drivers. To address this issue, this study proposes an online personalized preference learning method (OPPLM) that utilizes a pairwise comparison group preference query and the Bayesian approach. The proposed OPPLM adopts a two-layer hierarchical structure model based on utility theory to represent driver preferences on the trajectory. To improve the accuracy of learning, the uncertainty of driver query answers is modeled. In addition, informative query and greedy query selection methods are used to improve learning speed. To determine when the driver’s preferred trajectory has been found, a convergence criterion is proposed. To evaluate the effectiveness of the OPPLM, a user study is conducted to learn the driver’s preferred trajectory in the curve of the lane centering control (LCC) system. The results show that the OPPLM can converge quickly, requiring only about 11 queries on average. Moreover, it accurately learned the driver’s favorite trajectory, and the estimated utility of the driver preference model is highly consistent with the subject evaluation score.
- Open Access
Introspective False Negative Prediction for Black-Box Object Detectors in Autonomous Driving
Object detection plays a critical role in autonomous driving, but current state-of-the-art object detectors will inevitably fail in many driving scenes, which is unacceptable for safety-critical automated vehicles. Given the complexity of the real traffic scenarios, it is impractical to guarantee zero detection failure; thus, online failure prediction is of crucial importance to mitigate the risk of traffic accidents. Of all the failure cases, False Negative (FN) objects are most likely to cause catastrophic consequences, but little attention has been paid to the online FN prediction. In this paper, we propose a general introspection framework that can make online prediction of FN objects for black-box object detectors. In contrast to existing methods which rely on empirical assumptions or handcrafted features, we facilitate the FN feature extraction by an introspective FN predictor we designed in this framework. For this purpose, we extend the original concept of introspection to object-wise FN predictions, and propose a multi-branch cooperation mechanism to address the distinct foreground-background imbalance problem of FN objects. The effectiveness of the proposed framework is verified through extensive experiments and analysis, and the results show that our method successfully predicts the FN objects with 81.95% precision for 88.10% recall on the challenging KITTI Benchmark, and effectively improves object detection performance by taking FN predictions into consideration.
Open AccessIntrospective False Negative Prediction for Black-Box Object Detectors in Autonomous Driving
Object detection plays a critical role in autonomous driving, but current state-of-the-art object detectors will inevitably fail in many driving scenes, which is unacceptable for safety-critical automated vehicles. Given the complexity of the real traffic scenarios, it is impractical to guarantee zero detection failure; thus, online failure prediction is of crucial importance to mitigate the risk of traffic accidents. Of all the failure cases, False Negative (FN) objects are most likely to cause catastrophic consequences, but little attention has been paid to the online FN prediction. In this paper, we propose a general introspection framework that can make online prediction of FN objects for black-box object detectors. In contrast to existing methods which rely on empirical assumptions or handcrafted features, we facilitate the FN feature extraction by an introspective FN predictor we designed in this framework. For this purpose, we extend the original concept of introspection to object-wise FN predictions, and propose a multi-branch cooperation mechanism to address the distinct foreground-background imbalance problem of FN objects. The effectiveness of the proposed framework is verified through extensive experiments and analysis, and the results show that our method successfully predicts the FN objects with 81.95% precision for 88.10% recall on the challenging KITTI Benchmark, and effectively improves object detection performance by taking FN predictions into consideration.
- Open Access
Data Efficient Reinforcement Learning for Integrated Lateral Planning and Control in Automated Parking System
Reinforcement learning (RL) is a promising direction in automated parking systems (APSs), as integrating planning and tracking control using RL can potentially maximize the overall performance. However, commonly used model-free RL requires many interactions to achieve acceptable performance, and model-based RL in APS cannot continuously learn. In this paper, a data-efficient RL method is constructed to learn from data by use of a model-based method. The proposed method uses a truncated Monte Carlo tree search to evaluate parking states and select moves. Two artificial neural networks are trained to provide the search probability of each tree branch and the final reward for each state using self-trained data. The data efficiency is enhanced by weighting exploration with parking trajectory returns, an adaptive exploration scheme, and experience augmentation with imaginary rollouts. Without human demonstrations, a novel training pipeline is also used to train the initial action guidance network and the state value network. Compared with path planning and path-following methods, the proposed integrated method can flexibly co-ordinate the longitudinal and lateral motion to park a smaller parking space in one maneuver. Its adaptability to changes in the vehicle model is verified by joint Carsim and MATLAB simulation, demonstrating that the algorithm converges within a few iterations. Finally, experiments using a real vehicle platform are used to further verify the effectiveness of the proposed method. Compared with obtaining rewards using simulation, the proposed method achieves a better final parking attitude and success rate.
Open AccessData Efficient Reinforcement Learning for Integrated Lateral Planning and Control in Automated Parking System
Reinforcement learning (RL) is a promising direction in automated parking systems (APSs), as integrating planning and tracking control using RL can potentially maximize the overall performance. However, commonly used model-free RL requires many interactions to achieve acceptable performance, and model-based RL in APS cannot continuously learn. In this paper, a data-efficient RL method is constructed to learn from data by use of a model-based method. The proposed method uses a truncated Monte Carlo tree search to evaluate parking states and select moves. Two artificial neural networks are trained to provide the search probability of each tree branch and the final reward for each state using self-trained data. The data efficiency is enhanced by weighting exploration with parking trajectory returns, an adaptive exploration scheme, and experience augmentation with imaginary rollouts. Without human demonstrations, a novel training pipeline is also used to train the initial action guidance network and the state value network. Compared with path planning and path-following methods, the proposed integrated method can flexibly co-ordinate the longitudinal and lateral motion to park a smaller parking space in one maneuver. Its adaptability to changes in the vehicle model is verified by joint Carsim and MATLAB simulation, demonstrating that the algorithm converges within a few iterations. Finally, experiments using a real vehicle platform are used to further verify the effectiveness of the proposed method. Compared with obtaining rewards using simulation, the proposed method achieves a better final parking attitude and success rate.
- 5 August 2016
- Published by SAGE Publications in Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering
- Vol. 230 (12) , 1670-1686
A method to set the target on-centre steering force characteristic
In this paper, a method to set the target on-centre steering force characteristic is described. First, both the linear correlations and the non-linear correlations between subjective evaluations and objective metrics were established. During this process, subjective and objective evaluation tests regarding the on-centre steering force characteristic were conducted on a moving-base driving simulator by 15 experienced drivers. To ensure the reliability of the subjective ratings, a method to evaluate the ratings was introduced. Then the linear correlations and the non-linear correlations were analysed using multiple linear regression analysis and a neural network respectively. Second, by setting the desired subjective ratings as an optimization goal, target objective metrics were obtained on the basis of either the linear correlations or the non-linear correlations by means of non-linear programming. The target on-centre steering force characteristic was therefore designed on the basis of the target objective metrics. The results showed that the target on-centre steering force characteristic obtained from the linear correlations was quite similar to that obtained from the non-linear correlations. Also, it was verified that the target on-centre steering force characteristic matched the drivers’ preferences. From the correlations and the target on-centre steering force characteristic, engineers can establish the ratings of subjective evaluations and determine how to design the system to meet the target during the virtual design process. It shows a great potential in assisting the engineers to design a good steering feel at the early stages of the development process. Furthermore, the feasibility of this proposed method to allow customers to customize the on-centre steering force characteristic was explored. During this process, a fixed-base driving simulator was used instead of a moving-base driving simulator for cost reasons.
- 5 August 2016
- Published by SAGE Publications in Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering
- Vol. 230 (12) , 1670-1686
A method to set the target on-centre steering force characteristic
In this paper, a method to set the target on-centre steering force characteristic is described. First, both the linear correlations and the non-linear correlations between subjective evaluations and objective metrics were established. During this process, subjective and objective evaluation tests regarding the on-centre steering force characteristic were conducted on a moving-base driving simulator by 15 experienced drivers. To ensure the reliability of the subjective ratings, a method to evaluate the ratings was introduced. Then the linear correlations and the non-linear correlations were analysed using multiple linear regression analysis and a neural network respectively. Second, by setting the desired subjective ratings as an optimization goal, target objective metrics were obtained on the basis of either the linear correlations or the non-linear correlations by means of non-linear programming. The target on-centre steering force characteristic was therefore designed on the basis of the target objective metrics. The results showed that the target on-centre steering force characteristic obtained from the linear correlations was quite similar to that obtained from the non-linear correlations. Also, it was verified that the target on-centre steering force characteristic matched the drivers’ preferences. From the correlations and the target on-centre steering force characteristic, engineers can establish the ratings of subjective evaluations and determine how to design the system to meet the target during the virtual design process. It shows a great potential in assisting the engineers to design a good steering feel at the early stages of the development process. Furthermore, the feasibility of this proposed method to allow customers to customize the on-centre steering force characteristic was explored. During this process, a fixed-base driving simulator was used instead of a moving-base driving simulator for cost reasons.
- 1 December 2015
- Published by Springer Nature
- p. 631-638
Road Friction Coefficient Estimation by Using Lateral Dynamics Based on In-Wheel Motor Driven Electric Vehicle
- 1 December 2015
- Published by Springer Nature
- p. 631-638
Road Friction Coefficient Estimation by Using Lateral Dynamics Based on In-Wheel Motor Driven Electric Vehicle
- 23 October 2012
- Published by Springer Nature
- p. 313-327
Correlations Between Subjective and Objective Evaluations of On-Center Steering Feel
- 23 October 2012
- Published by Springer Nature
- p. 313-327
Correlations Between Subjective and Objective Evaluations of On-Center Steering Feel