Performance Metrics and Validation Methods for Vehicle Position Estimators

Fazekas, Adrian; Oeser, Markus

doi:10.1109/tits.2019.2920973

Cited by 4 publications

(3 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The second road was the test motorway used for the study, which was 4.5 km long and had an 800-m long entry lane at the beginning, where each participant started the simulation. The speed limit on the main road was 100 km/h, just like on the motorway section where we obtained real traffic data from videos with image processing methods ( 34 , 35 ). The entry lane had a speed limit of 80 km/h.…”

Section: Methodsmentioning

confidence: 99%

Car-Following Model Calibration Based on Driving Simulator Data to Study Driver Characteristics and to Investigate Model Validity in Extreme Traffic Situations

Berghaus

Kalló

Oeser

2021

Transportation Research Record: Journal of the Transportation R

Self Cite

View full text Add to dashboard Cite

In this paper we use traffic data from a driving simulator study to calibrate four different car-following models. We also present two applications for which the calibration results can be used. The first application relied on the advantage that driving simulator data also contain information on driver characteristics, for example, age, gender, or the self-assessment of driver behavior. By calibrating the models for each driver individually, the resulting model parameters could be used to analyze the influence of driver characteristics on driver behavior. The analysis revealed that certain characteristics, for example, self-identification as an aggressive driver, were reflected in the model parameters. The second application was based on the capability to simulate dangerous situations that require extreme driving behavior, which is often not included in datasets from real traffic and cannot be provoked in field studies. The model validity in these situations was analyzed by comparing the prediction errors of normal and extreme driving behavior. The results showed that all four car-following models underestimated the deceleration in an emergency braking scenario in which the drivers were momentarily shocked. The driving simulator study was validated by comparing the calibration results with those obtained from real trajectory data. We concluded that driving simulator data were suitable for the two proposed applications, although the validity of driving simulator studies must always be regarded.

show abstract

Section: Methodsmentioning

confidence: 99%

Car-Following Model Calibration Based on Driving Simulator Data to Study Driver Characteristics and to Investigate Model Validity in Extreme Traffic Situations

Berghaus

Kalló

Oeser

2021

Transportation Research Record: Journal of the Transportation R

Self Cite

View full text Add to dashboard Cite

show abstract

“…There are many contributions in literature with respect to object detection evaluation presented a multiplicity of metrics to measure different aspects of evaluation protocols [57]. Kasturi et al [58] presented metrics and tools for evaluating the performance of object detection algorithms based ( 16) on the detection accuracy (DA), multiple object detection accuracy (MODA), and multiple object detection precision (MODP).…”

Section: Evaluation Metrics and Datasetsmentioning

confidence: 99%

Local binary pattern-based on-road vehicle detection in urban traffic scene

Hassaballah

Kenk

El-Henawy

2020

Pattern Anal Applic

View full text Add to dashboard Cite

For intelligent traffic monitoring systems and related applications, detecting vehicles on roads is a vital step. However, robust and efficient vehicles detection is still a challenging problem due to variations in the appearance of the vehicles and complicated background of the roads. In this paper, we propose a simple and effective vehicle detection method based on local vehicle's texture and appearance histograms feed into clustering forests. The interdependency of vehicle's parts locations is incorporating within a clustering forests framework. Local binary pattern-like descriptors are utilized for texture feature extraction. Through utilizing the LBP descriptors, the local structures of vehicles, such as edge, contour and flat region can be effectively depicted. The align set of histograms generated concurrence with LBPs spatial for random sampled local regions are used to measure the dissimilarity between regions of all training images. Evaluating the fit between histograms is built in clustering forests. That is, clustering discriminative codebooks of latent features are used to search between different LBP features of the random regions utilizing the Chi-square dissimilarity measure. Besides, saliency maps built by the learnt latent features are adopted to determine the vehicles locations in test image. Effectiveness of the proposed method is evaluated on different car datasets stressing various imaging conditions and the obtained results show that the method achieves significant improvements compared to published methods.

show abstract

“…Most of these methods are to predict the macroscopic traffic flow, where they ignore the motions of individual vehicles. However, the traffic flow prediction is not helpful to vehicular networking [15] as networking is more about individuals instead of groups. In this context, we should study the vehicle trajectory prediction to target individual vehicles.…”

mentioning

confidence: 99%

A Novel Generation-Adversarial-Network-Based Vehicle Trajectory Prediction Method for Intelligent Vehicular Networks

Zhao

Liu

Al‐Dubai

et al. 2021

IEEE Internet Things J.

View full text Add to dashboard Cite

Prediction of the future location of vehicles and other mobile targets is instrumental in intelligent transportation system applications. In fact, networking schemes and protocols based on machine learning can benefit from the results of such accurate trajectory predictions. This is because routing decisions always need to be made for the future scenario due to the inevitable latency caused by processing and propagation of the routing request and response. Thus, to predict the highprecision trajectory beyond the state-of-the-art, we propose a Generative Adversarial Network-based VEhicle trajEctory Prediction method, GAN-VEEP, for urban roads. The proposed method consists of three components, 1) vehicle coordinate transformation for data set preparation, 2) neural network prediction model trained by GAN, and 3) vehicle turning model to adjust the prediction process. The vehicle coordinate transformation model is introduced to deal with the complex spatial dependence in the urban road topology. Then, the neural network prediction model learns from the behavior of vehicle drivers. Finally, the vehicle turning model can refine the driving path based on the driver's psychology. Compared with its counterparts, the experimental results show that GAN-VEEP exhibits higher effectiveness in terms of the Average Accuracy, Mean Absolute Error, and Root Mean Squared Error.

show abstract

Performance Metrics and Validation Methods for Vehicle Position Estimators

Cited by 4 publications

References 41 publications

Car-Following Model Calibration Based on Driving Simulator Data to Study Driver Characteristics and to Investigate Model Validity in Extreme Traffic Situations

Car-Following Model Calibration Based on Driving Simulator Data to Study Driver Characteristics and to Investigate Model Validity in Extreme Traffic Situations

Local binary pattern-based on-road vehicle detection in urban traffic scene

A Novel Generation-Adversarial-Network-Based Vehicle Trajectory Prediction Method for Intelligent Vehicular Networks

Contact Info

Product

Resources

About