Smart surveillance system for real-time multi-person multi-camera tracking at the edge

Gaikwad, Bipin; Karmakar, Arindam

doi:10.1007/s11554-020-01066-8

Cited by 21 publications

(2 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The current available MC-MPT approaches had a high level of computational complexity and were not adequately robust in the previously mentioned challenges. In [41], a real-time Attribute Recognition-based MC-MPT (AR-MC-MPT) framework was proposed that aimed to overcome these limitations. This framework employed an end-to-end approach for object detection, feature extraction, and attribute recognition.…”

Section: B Two-stage Approachmentioning

confidence: 99%

Learning to Track With Dynamic Message Passing Neural Network for Multi-Camera Multi-Object Tracking

Bilakeri,

Kotegar

2024

IEEE Access

View full text Add to dashboard Cite

Multi-camera multi-object tracking (MC-MOT) has become pivotal in various real-world applications within computer vision. Despite extensive research, solving the data association problem remains one of the most formidable challenges in MC-MOT pipeline. This challenge is compounded by factors such as varying illumination, diverse walking patterns, and trajectory occlusions. In recent years, graph neural networks (GNNs) have emerged as promising tools for enhancing data association. However, prevalent graph-based MC-MOT methods often rely on computationally inefficient min-cost flow approaches for cross-camera association, with static graph structures that lack adaptability to new detections. Moreover, these methods typically process cameras in pairs, leading to localized solutions rather than a holistic global approach. To address these limitations, we propose a two-stage lightweight crosscamera tracker designed to achieve a global solution efficiently. Our approach prioritizes the quality of local tracklets, enhancing them through supervised learning on multi-source datasets using the DeepSort model. For multi-camera association, we leverage the dynamic connectivity of Message Passing Graph Neural Networks (MPGNNs) to jointly learn features and similarities previously untapped in this domain. Our proposed model significantly improves detection accuracy and feature extraction, outperforming current MC-MOT algorithms on cross-camera benchmarks. This advancement marks a notable step forward in the field, offering more precise tracking capabilities and demonstrating the potential of integrating state-of-theart techniques for enhanced performance in complex tracking scenarios.

show abstract

Section: B Two-stage Approachmentioning

confidence: 99%

Learning to Track With Dynamic Message Passing Neural Network for Multi-Camera Multi-Object Tracking

Bilakeri,

Kotegar

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Video surveillance plays an emerging role for achieving intelligent transportation system for information extraction derived from the environment. The information extracted from real time scenario can be quite significant for detecting, classifying and summarization of abnormal events that includes investigations in video surveillance [1,2]. A number of researchers over the decades concentrated on the objective of detection, classification and summarization of abnormal events in reduced computational complexities and maximized event frame [3,4].…”

Section: Introductionmentioning

confidence: 99%

An integrated framework for abnormal event detection and video summarization using deep learning

2022

IJATEE

View full text Add to dashboard Cite

This field of research is considered to be essential for the enhancement of video surveillance systems that necessitates vast space of storage and analysis of complex data.Abnormal event detection, classification and video summarization is considered to be an important process that results in handling the challenges for improvement of video surveillance systems [5,6].The main issues in abnormal event detection and summarization is of immense concern to the research community in intelligent transportation systems [7,8]. It is very difficult from the state of the art techniques to provide the accurate detection and classification of abnormal events in the video frames andto summarize the key events. The core aim of the work is to develop the higher accuracy in detection, class ification and Re search Article

show abstract

Twenty-five years of real-time surveillance video analytics: a bibliometric review

Nallola,

Ayyasamy

2024

Multimed Tools Appl

View full text Add to dashboard Cite

Smart surveillance system for real-time multi-person multi-camera tracking at the edge

Cited by 21 publications

References 47 publications

Learning to Track With Dynamic Message Passing Neural Network for Multi-Camera Multi-Object Tracking

Learning to Track With Dynamic Message Passing Neural Network for Multi-Camera Multi-Object Tracking

An integrated framework for abnormal event detection and video summarization using deep learning

Twenty-five years of real-time surveillance video analytics: a bibliometric review

Contact Info

Product

Resources

About