arXiv:2511.13904v1 Announce Type: new 
Abstract: In modern Intelligent Transportation Systems (ITS), cameras are a key component due to their ability to provide valuable information for multiple stakeholders. A central task is Multi-Camera Vehicle Tracking (MCVT), which generates vehicle trajectories and enables applications such as anomaly detection, traffic density estimation, and suspect vehicle tracking. However, most existing studies on MCVT emphasize accuracy while overlooking real-time performance and scalability. These two aspects are essential for real-world deployment and become increasingly challenging in city-scale applications as the number of cameras grows. To address this issue, we propose SAE-MCVT, the first scalable real-time MCVT framework. The system includes several edge devices that interact with one central workstation separately. On the edge side, live RTSP video streams are serialized and processed through modules including object detection, object tracking, geo-mapping, and feature extraction. Only lightweight metadata -- vehicle locations and deep appearance features -- are transmitted to the central workstation. On the central side, cross-camera association is calculated under the constraint of spatial-temporal relations between adjacent cameras, which are learned through a self-supervised camera link model. Experiments on the RoundaboutHD dataset show that SAE-MCVT maintains real-time operation on 2K 15 FPS video streams and achieves an IDF1 score of 61.2. To the best of our knowledge, this is the first scalable real-time MCVT framework suitable for city-scale deployment.

إطار SAE-MCVT يمثل تقدمًا كبيرًا في أنظمة النقل الذكية (ITS) من خلال تقديم حل قابل للتوسع وفي الوقت الحقيقي لتتبع المركبات عبر عدة كاميرات (MCVT). يتناول هذا الإطار التحديات المتعلقة بالدقة والأداء في الوقت الحقيقي وقابلية التوسع، وهي أمور حاسمة للتطبيقات الحضرية التي تحتوي على عدد كبير من الكاميرات. من خلال استخدام الحوسبة على الحافة، يقوم SAE-MCVT بمعالجة تدفقات الفيديو المباشرة عبر وحدات متعددة، مما يتيح تطبيقات مثل تقدير كثافة المرور وتتبع المركبات المشتبه بها، وبالتالي تعزيز كفاءة أنظمة النقل.

El marco SAE-MCVT representa un avance significativa en los Sistemas de Transporte Inteligente (ITS) al ofrecer una solución escalable y en tiempo real para el Seguimiento de Vehículos Multi-Cámara (MCVT). Este marco aborda los desafíos de precisión, rendimiento en tiempo real y escalabilidad, que son cruciales para aplicaciones urbanas con numerosas cámaras. Al utilizar la computación en el borde, SAE-MCVT procesa flujos de video en vivo a través de varios módulos, permitiendo aplicaciones como la estimación de la densidad del tráfico y el seguimiento de vehículos sospechosos, mejorando así l…

Le cadre SAE-MCVT représente une avancée significative dans les systèmes de transport intelligents (ITS) en offrant une solution évolutive et en temps réel pour le suivi de véhicules multi-caméras (MCVT). Ce cadre répond aux défis d'exactitude, de performance en temps réel et d'évolutivité, qui sont cruciaux pour les applications urbaines avec de nombreuses caméras. En utilisant l'informatique en périphérie, SAE-MCVT traite des flux vidéo en direct à travers divers modules, permettant des applications telles que l'estimation de la densité de trafic et le suivi de véhicules suspects, améliorant…

The SAE-MCVT framework represents a significant advancement in Intelligent Transportation Systems (ITS) by offering a scalable and real-time solution for Multi-Camera Vehicle Tracking (MCVT). This framework addresses the challenges of accuracy, real-time performance, and scalability, which are crucial for urban applications with numerous cameras. By utilizing edge computing, SAE-MCVT processes live video streams through various modules, enabling applications such as traffic density estimation and suspect vehicle tracking, thereby enhancing the efficiency of transportation systems.

SAE-MCVT: A Real-Time and Scalable Multi-Camera Vehicle Tracking Framework Powered by Edge Computing

arXiv:2511.16929v1 Announce Type: new 
Abstract: Detecting trajectory anomalies is a vital task in modern Intelligent Transportation Systems (ITS), enabling the identification of unsafe, inefficient, or irregular travel behaviours. While deep learning has emerged as the dominant approach, several key challenges remain unresolved. First, sub-trajectory anomaly detection, capable of pinpointing the precise segments where anomalies occur, remains underexplored compared to whole-trajectory analysis. Second, many existing methods depend on carefully tuned thresholds, limiting their adaptability in real-world applications. Moreover, the irregular sampling of trajectory data and the presence of noise in training sets further degrade model performance, making it difficult to learn reliable representations of normal routes. To address these challenges, we propose a contrastive reinforcement learning framework for online trajectory anomaly detection, CroTad. Our method is threshold-free and robust to noisy, irregularly sampled data. By incorporating contrastive learning, CroTad learns to extract diverse normal travel patterns for different itineraries and effectively distinguish anomalous behaviours at both sub-trajectory and point levels. The detection module leverages deep reinforcement learning to perform online, real-time anomaly scoring, enabling timely and fine-grained identification of abnormal segments. Extensive experiments on two real-world datasets demonstrate the effectiveness and robustness of our framework across various evaluation scenarios.

تم تقديم إطار جديد يسمى CroTad للكشف عن الشذوذ في المسارات عبر الإنترنت، مما يعالج التحديات الحرجة في أنظمة النقل الذكية (ITS). تستخدم هذه الطريقة التعلم المعزز التبايني للكشف عن الشذوذ في المسارات الفرعية دون الاعتماد على عتبات محددة مسبقًا، مما يعزز القدرة على التكيف في التطبيقات الواقعية.

Se ha presentado un nuevo marco llamado CroTad para la detección de anomalías en trayectorias en línea, abordando desafíos críticos en los Sistemas de Transporte Inteligente (ITS). Este método utiliza el aprendizaje por refuerzo contrastivo para detectar anomalías en subtrayectorias sin depender de umbrales predefinidos, mejorando así la adaptabilidad en aplicaciones del mundo real.

Un nouveau cadre nommé CroTad a été introduit pour la détection des anomalies de trajectoire en ligne, abordant des défis critiques dans les systèmes de transport intelligents (ITS). Cette méthode utilise l'apprentissage par renforcement contrastif pour détecter les anomalies dans les sous-trajectoires sans dépendre de seuils prédéfinis, améliorant ainsi l'adaptabilité dans les applications réelles.

A new framework named CroTad has been introduced for online trajectory anomaly detection, addressing critical challenges in Intelligent Transportation Systems (ITS). This method utilizes contrastive reinforcement learning to detect anomalies in sub-trajectories without relying on predefined thresholds, enhancing adaptability in real-world applications.

SAE-MCVT: A Real-Time and Scalable Multi-Camera Vehicle Tracking Framework Powered by Edge Computing

Was this article worth reading? Share it

Velory

Snapshot AI

MyCVBot