Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
6.8
3.9
Journals
Sensors
Special Issues
Sensor-Based Object Detection and Recognition in Intelligent Surveillance Systems
sensors-logo
Submit to Sensors Review for Sensors Propose a Special Issue

Journal Menu

► Journal Menu

Journal Browser

► Journal Browser
share announcement

Sensor-Based Object Detection and Recognition in Intelligent Surveillance Systems

A special issue of Sensors (ISSN 1424-8220). This special issue belongs to the section "Intelligent Sensors".

Deadline for manuscript submissions: closed (20 January 2024) | Viewed by 7596

Special Issue Editor

Dr. Amir Atapour-Abarghouei

E-Mail Website
Guest Editor
Department of Computer Science, Durham University, Durham DH1 3LE, UK
Interests: machine learning; computer vision; 3D scene analysis; semantic and geometric scene understanding; scene depth prediction

Special Issue Information

Dear Colleagues,

Recent advances in modern object detection and recognition have enabled improvements in security and surveillance applications, and have supported the automation of numerous crime prevention and safety procedures. The use of these surveillance systems, often powered by artificial intelligence and computer vision, requires that the reliability and safeguarding of such systems is ensured in order to prevent any malicious intervention from bad actors attempting to circumvent any security procedures in place. This Special Issue is motivated by the synergetic relationships between surveillance systems taking advantage of object detection, recognition and anomaly detection, as well as the need to combat algorithms and actors attempting to attack such systems.

We encourage submissions from all areas of computer vision, focusing on the use of scene understanding techniques including object detection, recognition and anomaly detection and their reliability. More general contributions such as novel theories, frameworks, architectures, and datasets are also welcome. The topics of interest include, but are not limited to, the following:

  • Object detection and recognition;
  • Anomaly detection for surveillance;
  • Detecting temporal abnormalities;
  • Object-level anomaly detection;
  • Threat detection and prediction;
  • Person re-identification;
  • Thermal imaging for surveillance;
  • Action and gesture recognition;
  • Gait recognition and analysis;
  • Two/three-dimensional human pose estimation;
  • Face detection, recognition, and verification;
  • Virtual fencing;
  • Sensitive site/infrastructure security;
  • Object detection and anomaly detection for aerial surveillance;
  • Traffic infraction/incident detection;
  • COVID-19 compliance;
  • Object detection in panoramic and 360° images;
  • Adversarial examples;
  • Black box and white box attacks;
  • Adversarial example detection;
  • Morphing attack detection;
  • Preventing evasion and poison attacks;
  • Fair AI for security and surveillance;
  • Bias identification and removal in surveillance systems;
  • Combatting data imbalance in object detection;
  • Alternative methods for the security of surveillance systems.

Dr. Amir Atapour-Abarghouei
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Sensors is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • surveillance technologies
  • object detection
  • object recognition
  • anomaly detection
  • thermal sensing
  • panoramic and 360° imaging
  • adversarial examples
  • bias identification and removal

Published Papers (3 papers)

Download All Papers
Order results
Result details
Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:

Research

22 pages, 8952 KiB  
Article
On the Importance of Attention and Augmentations for Hypothesis Transfer in Domain Adaptation and Generalization
by Rajat Sahay, Georgi Thomas, Chowdhury Sadman Jahan, Mihir Manjrekar, Dan Popp and Andreas Savakis
Sensors 2023, 23(20), 8409; https://doi.org/10.3390/s23208409 - 12 Oct 2023
Cited by 1 | Viewed by 1197
Abstract
Unsupervised domain adaptation (UDA) aims to mitigate the performance drop due to the distribution shift between the training and testing datasets. UDA methods have achieved performance gains for models trained on a source domain with labeled data to a target domain with only [...] Read more.
Unsupervised domain adaptation (UDA) aims to mitigate the performance drop due to the distribution shift between the training and testing datasets. UDA methods have achieved performance gains for models trained on a source domain with labeled data to a target domain with only unlabeled data. The standard feature extraction method in domain adaptation has been convolutional neural networks (CNNs). Recently, attention-based transformer models have emerged as effective alternatives for computer vision tasks. In this paper, we benchmark three attention-based architectures, specifically vision transformer (ViT), shifted window transformer (SWIN), and dual attention vision transformer (DAViT), against convolutional architectures ResNet, HRNet and attention-based ConvNext, to assess the performance of different backbones for domain generalization and adaptation. We incorporate these backbone architectures as feature extractors in the source hypothesis transfer (SHOT) framework for UDA. SHOT leverages the knowledge learned in the source domain to align the image features of unlabeled target data in the absence of source domain data, using self-supervised deep feature clustering and self-training. We analyze the generalization and adaptation performance of these models on standard UDA datasets and aerial UDA datasets. In addition, we modernize the training procedure commonly seen in UDA tasks by adding image augmentation techniques to help models generate richer features. Our results show that ConvNext and SWIN offer the best performance, indicating that the attention mechanism is very beneficial for domain generalization and adaptation with both transformer and convolutional architectures. Our ablation study shows that our modernized training recipe, within the SHOT framework, significantly boosts performance on aerial datasets. Full article
(This article belongs to the Special Issue Sensor-Based Object Detection and Recognition in Intelligent Surveillance Systems)
►▼ Show Figures

Figure 1

25 pages, 1602 KiB  
Article
CNN-ViT Supported Weakly-Supervised Video Segment Level Anomaly Detection
by Md. Haidar Sharif, Lei Jiao and Christian W. Omlin
Sensors 2023, 23(18), 7734; https://doi.org/10.3390/s23187734 - 07 Sep 2023
Cited by 2 | Viewed by 1274
Abstract
Video anomaly event detection (VAED) is one of the key technologies in computer vision for smart surveillance systems. With the advent of deep learning, contemporary advances in VAED have achieved substantial success. Recently, weakly supervised VAED (WVAED) has become a popular VAED technical [...] Read more.
Video anomaly event detection (VAED) is one of the key technologies in computer vision for smart surveillance systems. With the advent of deep learning, contemporary advances in VAED have achieved substantial success. Recently, weakly supervised VAED (WVAED) has become a popular VAED technical route of research. WVAED methods do not depend on a supplementary self-supervised substitute task, yet they can assess anomaly scores straightway. However, the performance of WVAED methods depends on pretrained feature extractors. In this paper, we first address taking advantage of two pretrained feature extractors for CNN (e.g., C3D and I3D) and ViT (e.g., CLIP), for effectively extracting discerning representations. We then consider long-range and short-range temporal dependencies and put forward video snippets of interest by leveraging our proposed temporal self-attention network (TSAN). We design a multiple instance learning (MIL)-based generalized architecture named CNN-ViT-TSAN, by using CNN- and/or ViT-extracted features and TSAN to specify a series of models for the WVAED problem. Experimental results on publicly available popular crowd datasets demonstrated the effectiveness of our CNN-ViT-TSAN. Full article
(This article belongs to the Special Issue Sensor-Based Object Detection and Recognition in Intelligent Surveillance Systems)
►▼ Show Figures

Figure 1

14 pages, 42091 KiB  
Article
POSEIDON: A Data Augmentation Tool for Small Object Detection Datasets in Maritime Environments
by Pablo Ruiz-Ponce, David Ortiz-Perez, Jose Garcia-Rodriguez and Benjamin Kiefer
Sensors 2023, 23(7), 3691; https://doi.org/10.3390/s23073691 - 02 Apr 2023
Cited by 12 | Viewed by 4557
Abstract
Certain fields present significant challenges when attempting to train complex Deep Learning architectures, particularly when the available datasets are limited and imbalanced. Real-time object detection in maritime environments using aerial images is a notable example. Although SeaDronesSee is the most extensive and complete [...] Read more.
Certain fields present significant challenges when attempting to train complex Deep Learning architectures, particularly when the available datasets are limited and imbalanced. Real-time object detection in maritime environments using aerial images is a notable example. Although SeaDronesSee is the most extensive and complete dataset for this task, it suffers from significant class imbalance. To address this issue, we present POSEIDON, a data augmentation tool specifically designed for object detection datasets. Our approach generates new training samples by combining objects and samples from the original training set while utilizing the image metadata to make informed decisions. We evaluate our method using YOLOv5 and YOLOv8 and demonstrate its superiority over other balancing techniques, such as error weighting, by an overall improvement of 2.33% and 4.6%, respectively. Full article
(This article belongs to the Special Issue Sensor-Based Object Detection and Recognition in Intelligent Surveillance Systems)
►▼ Show Figures

Figure 1

Show export options expand_more Show export options expand_less
Select all
Export citation of selected articles as:
 
Displaying articles 1-3
Back to TopTop