A Two-Phase Anomaly Detection Model for Secure Intelligent Transportation Ride-Hailing Trajectories

Belhadi, Asma; Djenouri, Youcef; Srivastava, Gautam; Djenouri, Djamel; Cano, Alberto; Lin, Jerry Chun‐Wei

doi:10.1109/tits.2020.3022612

Cited by 47 publications

(13 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The method in this paper uses the Kalman filter to predict the next position and bounding box size of the current basketball target position and obtains the trajectory prediction result of the basketball target. Similarity completes the tracking of multiple targets of the same category and corrects the problem of target occlusion in the tracking process and the problem of abnormal trajectory tracking caused by target occlusion, so the correction effect on abnormal target trajectories is better [21,22].…”

Section: Trajectory Correction Resultsmentioning

confidence: 99%

Multi-target trajectory tracking in multi-frame video images of basketball sports based on deep learning

Gong

Srivastava

2022

EAI Endorsed Scal Inf Syst

View full text Add to dashboard Cite

INTRODUCTION: There is occlusion interference in the multi-target visual tracking process of basketball video images, which leads to poor accuracy of multi-target trajectory tracking. This paper studies the multi-target trajectory tracking method in multi-frame video images of basketball sports based on deep learning. OBJECTIVES: Aiming at the problem of target occlusion in the tracking process and the problem of trajectory tracking anomaly caused by target occlusion, a modified algorithm is proposed. METHODS: The method is divided into two parts: detection and tracking. In the detection part, the YOLOv3 algorithm in deep learning technology is used to detect each target in the video, and the original YOLOv3 backbone network Darknet-53 is replaced by the lightweight backbone network MobileNetV2 to extract the target features. RESULTS: Based on the target detection results, the Kalman filter is used to predict the next position and bounding box size of the target to obtain the target trajectory prediction results according to the current target position, then a hierarchical data association algorithm is designed, and multi-target tracking of the same category is completed based on the target appearance feature similarity and feature similarity. CONCLUSION: The experimental results show that the method can accurately detect the targets in multi-frame video images in basketball sports and obtain high-precision target trajectory tracking results.

show abstract

Section: Trajectory Correction Resultsmentioning

confidence: 99%

Multi-target trajectory tracking in multi-frame video images of basketball sports based on deep learning

Gong

Srivastava

2022

EAI Endorsed Scal Inf Syst

View full text Add to dashboard Cite

show abstract

“…If these kinds of data are provided, we could also match the event and taxi travel demand changing rule. This matching work could help to detect anomalies according to taxi data, such as indicated in the work of Davis et al [ 37 ] and Belhadi et al [ 38 ]. Moreover, in some situations the degree of dependency depends also on the operational conditions of the travel system.…”

Section: Discussionmentioning

confidence: 99%

A Study on the Decay Model of Multi-Block Taxi Travel Demand under the Influence of Major Urban Public Health Events

Luo

Huang

Zheng

2022

IJERPH

View full text Add to dashboard Cite

A sudden major public health event is likely to have a negative impact on public transport travel for residents, with public travel modes such as the metro and conventional buses experiencing varying degrees of decline in patronage. As a complement to public transport, taxi travel will suffer the same impact. Land use and population density among various functional blocks in a city are different, and therefore their changing rates in taxi travel demand are varied. This paper reveals the taxi travel demand correlations between urban blocks and then constructs a taxi travel demand decay model based on the Dynamic Input-Output Inoperability Model (DIIM) to simulate the decay degree of taxi travel demand in each block. When a major public health event occurs, the residential panic levels in different functional blocks may vary. It results in variable changing speeds of residential travel demand in each block. Based on this assumption, we use the intensity of travel demand as a correlation strength factor between blocks, and equate it with the technical coefficient in the DIIM model. We also define other variables to serve in model construction. These variables include the decay degree of travel demand intensity, residential travel willingness, coefficient of travel demand decay, derivative coefficient of travel demand interdependency, and demand perturbation coefficient. Lastly, we select a central area of Ningbo as the study area, and use taxi travel data in Ningbo during the COVID-19 pandemic of 2020 as input, simulate taxi travel demand dynamics, and analyze the accuracy and sensitivity of the model parameters. The relative errors between the five types of blocks and the actual decay of travel demand intensity are 8.3%, 3.8%, 8.7%, 5.5%, and 5.3%, respectively, which can basically match the actual situation, proving the validity of the model. The results of the study reveal the pattern of taxi travel demand decay among various blocks after major public health events. It provides methodological reference for decision makers to understand the development trend of multi-block taxi travel demand, so as to help form effective emergency plans for different blocks.

show abstract

“…It investigated the various correlations between the sequence data and identified the frequent patterns to create and train the isolation forest structure. In the intelligent transportation context, Belhadi et al [22] suggested a twophase secure-based method for identifying abnormalities from ride-hailing trajectories. The first phase seeks to identify taxi fraud by computing the distance between each stop point in each taxi route, while the second seeks to enhance the mining process via the use of both feature selection and sliding windows techniques.…”

Section: Related Workmentioning

confidence: 99%

“…Table I presents the merit and the limitation of the relevant works to this research study. Simple, contextual, collective outliers Use traditional techniques Yu et al [13] Both anomaly detection and forecasting Hard to estimate the time window Yamanishi et al [15] Use incremental learning Hard to build the training data Xie et al [16] Identify different kind of outliers Use traditional techniques Nesa et al [17] Detecting Outliers from IoT data Hard to build the training data Na et al [18] Less memory consumption Use tradition techniques Zhang et al [20] Use hybrid deep learning models Hard to build the training data Feremans et al [21] Study pattern correlation High time consuming Belhadi et al [22] Application to taxi frauds High time consuming Javed et al, [23] Use hybrid deep learning models Hard to build the training data Wang et al [24] Detecting anomaly sensors Use traditional techniques…”

Section: Related Workmentioning

confidence: 99%

Hybrid Group Anomaly Detection for Sequence Data: Application to Trajectory Data Analytics

Belhadi

Djenouri

Srivastava

et al. 2022

IEEE Trans. Intell. Transport. Syst.

Self Cite

View full text Add to dashboard Cite

Many research areas depend on group anomaly detection. The use of group anomaly detection can maintain and provide security and privacy to the data involved. This research attempts to solve the deficiency of the existing literature in outlier detection thus a novel hybrid framework to identify group anomaly detection from sequence data is proposed in this paper. It proposes two approaches for efficiently solving this problem: i) Hybrid Data Mining-based algorithm, consists of three main phases: first, the clustering algorithm is applied to derive the micro-clusters. Second, the kN N algorithm is applied to each micro-cluster to calculate the candidates of the group's outliers. Third, a pattern mining framework gets applied to the candidates of the group's outliers as a pruning strategy, to generate the groups of outliers, and ii) a GPU-based approach is presented, which benefits from the massively GPU computing to boost the runtime of the hybrid data mining-based algorithm. Extensive experiments were conducted to show the advantages of different sequence databases of our proposed model. Results clearly show the efficiency of a GPU direction when directly compared to a sequential approach by reaching a speedup of 451. In addition, both approaches outperform the baseline methods for group detection.

show abstract

A Two-Phase Anomaly Detection Model for Secure Intelligent Transportation Ride-Hailing Trajectories

Cited by 47 publications

References 47 publications

Multi-target trajectory tracking in multi-frame video images of basketball sports based on deep learning

Multi-target trajectory tracking in multi-frame video images of basketball sports based on deep learning

A Study on the Decay Model of Multi-Block Taxi Travel Demand under the Influence of Major Urban Public Health Events

Hybrid Group Anomaly Detection for Sequence Data: Application to Trajectory Data Analytics

Contact Info

Product

Resources

About