Smart Camera Research Articles

Secure Search of Smart Home Data Based on Access Control Encryption

Abstract: In recent years, the increasing variety of smart home devices, such as smart locks, surveillance cameras, and various sensors, has led to a significant rise in related data. To facilitate the sharing of smart home data, manufacturers store these data on the Internet of Things (IoT). However, since IoT servers are not fully trustworthy and the data often contain sensitive commercial information, any leakage could result in a commercial crisis for manufacturers. To address security concerns, manufacturers encrypt the data before storing them on the IoT. A challenge arises: how can users search for target data in encrypted form? This paper proposes a new search encryption scheme for smart home data, addressing the limitations of existing attribute-based encryption (ABE) searchable encryption schemes. The proposed scheme integrates access control trees and bilinear mapping technology, supporting multiple access policies including AND gates, OR gates, and threshold gates. Experimental results demonstrate a comprehensive analysis of the attribute-based searchable encryption protocol under a multi-authority architecture, covering four dimensions: correctness, security, complexity, and real-world application scenarios. In specific scenarios, when a data user, out of privacy concerns, is unwilling to reveal all attribute information to a single authority during a query for shared datathus avoiding identity disclosureor when the users identity information is managed by multiple independent authorities, this protocol proves highly effective. It ensures smooth data sharing while maintaining data security for both data owners and users, fully safeguarding data privacy and shareability.

The Future of Smart Home Security: Generative AI and LLMs for Intelligent Event Detection and Personalized Notifications

Abstract—Smart home security cameras are becoming more common, but their usefulness can be diminished by notification fatigue from too many alerts about minor incidents. This paper examines the gaps of existing event detection and notification systems in security cameras and then recommends using Generative AI and Large Language Models (LLMs) to add intelligence which would improve user experience. Generative AI can be leveraged to classify events more accurately and assist with anomaly detection. LLMs can further be used to create notifications that are tailored to the context and personalized to users behavior, helping to reduce notification fatigue and provide meaningful user alerts. The paper also looks into wider applications of these technologies to add intelligence and improve other related experiences like automated video summarization, proactive security measures, and improved privacy controls. The integration of Generative AI and LLMs with smart home security camera systems advances the smart cameras capabilities and offers enhanced security, personalized user experiences. Keywords—Smart home security, Generative AI, Large Language Models (LLMs), Event detection, Anomaly detection, Notification fatigue, Context-aware notifications, Personalized security, Reinforcement Learning from Human Feedback (RLHF), Internet of Things (IoT).

“They Didn’t Buy Their Smart TV to Watch Me with the Kids”: Comparing Nannies’ and Parents’ Privacy Threat Models for Smart Home Devices

Smart home devices raise privacy concerns among not only primary users, but also bystanders like domestic workers. We conducted 25 qualitative interviews with nannies and 16 with parents who employed nannies, in the US, to explore and compare their views on and privacy threat models for smart home devices. We found device-specific purposes of use inspired different perspectives among nanny participants. Most were comfortable with employers’ smart speakers and smart TVs, whose purpose had nothing to do with them. However, with indoor smart cameras, nanny participants were often not just bystanders but targets of monitoring; in such situations, they had a wider range of attitudes. In contrast, parent participants tended to have more similar views across devices. We found notable disconnects regarding disclosure, where nanny participants often hesitated to ask about cameras, but parent participants assumed nannies just didn’t care. We recommend prioritizing interventions supporting disclosure, discussion, and sharing control.

Facile fabrication of hemicrystalline bio-based polycarbonate with superhydrophobic and high transmittance

Developing materials that combine superhydrophobic properties with high optical transmittance poses a significant challenge. In this study, hydroxyl-terminated polydimethylsiloxane (HT-PDMS) and isosorbide (ISB) were polymerized in a single step to create a covalently bonded optical polycarbonate material. By inducing the formation of micro-papillary and lotus-shaped nanoscale structures via a solvent-triggered process, we significantly enhanced the “air cushion” effect, achieving a structure scale of approximately 5–7 μm. This resulted in a water contact angle of 157° while maintaining over 90 % optical transmittance. The structures were uniformly distributed throughout the polymer matrix, leading to a 500 % increase in tensile strength at break compared to pure isosorbide polycarbonate, with a maximum strength exceeding 50 MPa. These multifunctional materials show great promise for applications in smart windows, solar panels, camera lenses, and other optoelectronic devices.

IoT Based Smart Traffic Management System using Image Processing and Automated Street Lighting System

Through the integration of image processing, IoT, and automatic street lighting, this study presents a novel solution to urban traffic management. Real-time traffic data, such as vehicle counts and congestion levels, are continuously collected by smart cameras positioned at strategic intersections. A central control unit receives this data and uses sophisticated algorithms to dynamically modify traffic signal timings. Additionally, by including sensors to identify surrounding cars and pedestrians, an automatic street lighting system improves sustainability. This system helps save energy and money by optimizing energy consumption through the ability to change lighting intensity levels as needed. In general, this Internet of Things (IoT)-based smart traffic management system improves efficiency and safety for both commuters and locals by addressing congestion and fostering sustainable urban development

ВЛИЯНИЕ АДАПТИВНОГО УПРАВЛЕНИЯ СВЕТОФОРНЫМИ ОБЪЕКТАМИ НА ХАРАКТЕРИСТИКИ ДОРОЖНОГО ДВИЖЕНИЯ СМЕЖНЫХ ПЕРЕКРЕСТКОВ

The article is devoted to the study of the impact of adaptive algorithms for controlling traffic lights at one intersection on changes in traffic at adjacent intersections. The main purpose of the study is to identify the advantages and disadvantages of implementing such algorithms to assess the impact on adjacent intersections. Three algorithms have been proposed for the study: queue skipping, phase duration redistribution, and waiting time estimation. The study included traffic modeling, data collection on waiting times, queue lengths, and an assessment of changes in traffic flows. Data was collected using smart cameras, which allowed us to identify how changing traffic light work plans at one intersection affects the dynamics of traffic at neighboring intersections. Possible changes in traffic at adjacent intersections are proved, in particular, an increase in the number of waiting vehicles and the average waiting time at adjacent intersections, which indicates the negative impact of adaptive algorithms on adjacent intersections.

Forensically analyzing IoT smart camera using MAoIDFF-IoT framework

IoT devices spread over a wide range of applications these days, and their vast amount of data generated becomes a target for intruders. IoT digital forensics, which involves extracting the digital evidence from the IoT device itself and/or its network traffic using a framework is important and challenging. The challenges include the diversity of types of IoT devices, resource constraints, and users’ privacy. In this article, we focus on network forensics investigations of smart camera traffic as a case study. The investigation process followed the MAoIDFF-IoT framework, a comprehensive and effective framework for IoT devices, and focusing on the locations of potential Artifacts of Interest (AoI). In addition, a few scenarios in using the camera are investigated to obtain a valuable artifact. The results show that it is possible to extract a few artifacts from the network captured traffic even though the traffic is encrypted. Moreover, this research offers guidelines for digital investigators to conduct network forensics on smart camera devices, with detailed results provided.

A novel approach to classify pharyngitis using potential features based on doctor’s perspective

Pharyngitis is an inflammation of the oropharynx’s mucous membranes. It is typically brought on by a bacterial illness. The outburst of latest technologies has created the need for remote care of detecting diseases like pharyngitis through images of throat taken with help of smart camera. In recent years, research has forwarded with help of deep learning in classifying pharyngitis. But deep learning models require at least one hour training and requires considerably large data set to get a good accuracy. In this paper, we focused on this time constraint and are proposing a novel approach PFDP to classify pharyngitis through detection of potential features based on doctor’s perspective. We have extracted the tiny portions of image which the doctor observes them as infected and calculated frequencies of the occurrences of these portions and are given to custom made decision rules. The classification results showed significant improvement in performance in terms of time taken to reach average accuracy of 70%. It has taken only 5 minutes to extract counts of infected patterns and 1 more minute to get classification results by decision rules of if-then-else rules. We have conducted the experiment on set of 800 images. Though accuracy is lesser than that of what other works achieved but time taken to extract features is significantly lower than that of previous works. Also our approach does not require training and can be applied where scarcity of dataset exists. We assure that our approach is a new direction of research and can compete with more state of the art works in future.

Reliable and efficient integration of AI into camera traps for smart wildlife monitoring based on continual learning

In this paper, we comprehensively report on an efficient approach for the integration of artificial intelligence (AI) processing pipelines in camera traps for smart on-site wildlife monitoring. Our work covers hardware, software, and algorithmics. We have built two prototypes of smart camera trap on a maximum bill of materials of 100$. We have also built two datasets, made publicly available, comprising over 17 k images, many of them notably challenging even for humans. Leveraging our broad expertise on embedded systems, specialized software libraries and toolchains, and AI techniques such as transfer learning, explainable AI, and, most importantly, continual learning, we achieve more reliable inference on-site - specifically 10 % higher F1-score - than MegaDetector run off-site on a desktop computer. The paper includes many practical details on system realization and on-site training in addition to a vast set of lab and experimental results.

Security Evaluation of Companion Android Applications in IoT: The Case of Smart Security Devices.

Smart security devices, such as smart locks, smart cameras, and smart intruder alarms are increasingly popular with users due to the enhanced convenience and new features that they offer. A significant part of this convenience is provided by the device's companion smartphone app. Information on whether secure and ethical development practices have been used in the creation of these applications is unavailable to the end user. As this work shows, this means that users are impacted both by potential third-party attackers that aim to compromise their device, and more subtle threats introduced by developers, who may track their use of their devices and illegally collect data that violate users' privacy. Our results suggest that users of every application tested are susceptible to at least one potential commonly found vulnerability regardless of whether their device is offered by a known brand name or a lesser-known manufacturer. We present an overview of the most common vulnerabilities found in the scanned code and discuss the shortcomings of state-of-the-art automated scanners when looking at less structured programming languages such as C and C++. Finally, we also discuss potential methods for mitigation, and provide recommendations for developers to follow with respect to secure coding practices.

Smart Cities Driven by IoT: Improving Attack Detection with Cloud-Enabled Cybersecurity

The susceptibility of smart equipment to sophisticated cyber threats, such as Distributed Denial of Service (DDoS) assaults, makes anomaly detection in smart cities essential in the current era of information technology. A machine learning-driven method is presented to tackle this problem, which makes use of the power consumption patterns of Internet of Things (IoT) devices to spot unusual activity in smart home settings. The training model is implemented near IoT layer devices by utilizing a distributed fog network. To gather statistics on power utilization during daily activities, a Raspberry Pi prototype smart camera has been constructed. To produce power consumption patterns suggestive of DDoS assaults, the study also investigates anomaly identification using Artificial Neural Network (ANN) approaches. ANNs are used to simulate DDoS attacks on an experimental setup. With an average detection accuracy of 78.67% across a range of assault situations, the results demonstrate that a deep feed-forward neural network model works better than previous models. This strategy puts less demand on Internet of Things devices with limited resources, which qualifies it for use in smart city applications.

Towards the identification of transmission pathways and early detection of Enterococcus cecorum infection in broiler chickens

Enterococcus cecorum (EC) infection is an emerging endemic disease in UK and global broiler poultry with major economic impact and welfare concerns. There are significant research gaps with regards to EC pathogenesis, source of infection, transmission routes and early detection of disease, which this study aimed to address. In this prospective study, 725 environmental samples were collected from 4 broiler farms (A-D) the day before chick placement (d 1) and through the subsequent crop (d 7, 14, and 21). Cecal swabs were collected from birds that died of natural causes during the study period. A sample of birds that had been found dead or were culled for health reasons, were presented for post-mortem and samples were taken from lesions for EC culture. DNA was extracted from all environmental samples and EC detected using a qPCR and MALDI-TOF. Two EC isolates from diseased birds were inoculated on concrete slabs and incubated at 23°C and 32°C followed by swabbing of concrete culturing and determination of EC cfu at defined time points. Alongside environmental and bird sampling commercially available, smart camera systems were installed in selected houses on each farm to monitor bird activity and distribution. No EC outbreak occurred during the study, however, it was detected by qPCR in 215/725 (29.7 %) of all samples collected. Also, EC DNA was detected on average in 37% of samples collected on d 1, with approx. 88% of samples from chick paper being positive. Despite this, it was only cultured from 3 ceca samples and joint fluids of two infected birds from farm B on d 14 and 21. The survival experiments using isolates from infected chickens showed EC can survive on concrete for at least 21 d. This study provides invaluable insights into transmission pathways and tenacity of EC. Further studies are needed to determine strain characteristics in relation to their ability to cause disease and to further elucidate the sources of infection on poultry farms.

A detection and defense security system design for nuclear waste storage against stealth terrorists attack

Abstract The physical protection system (PPS) is one of the main systems used in nuclear facilities to protect these facilities from terrorist attacks, the possibility of bombing, sabotage, and theft. Installation of a security system for nuclear waste stores from the outside is necessary and very important for early detection and physiological deterrence greater than the internal security system. From this perspective, this work provides a solution for the protection of nuclear waste stores, which means detection and defense systems. This paper introduces an advanced security system and proposes a design for a detection and defense security system for nuclear waste storage against stealth terrorist attacks. The security system that is being suggested will have smart camera surveillance system, electric chock system, intense flashlight barrier, high-intensity sound barriers, and dual-side infrared rays. Three consecutive processes – detection, deterrence, and defense – will be operated around the building of the waste storage. The concept may be implemented by establishing an early detection system to detect any intrusion by terrorist attackers through a protected area surrounded by storage. The following step is deterring the attackers from completing their mission by installing advanced deterrence system and putting delay components along the path of their attack on the waste storage to complicate the mission. Then the defense system is operating to stop terrorist progress. Finally, the work make an evaluation process of the overall security system which has been designed by using EASI program to verify the accepted level of physical protection, maintain the validity and updates of security system components, and follow procedures, rules and requirements of the nuclear regulatory authority at the national level and meet the international IAEA concept and recommendations.

A proposed design for security system for nuclear waste warehouses against terrorists attack

A physical protection system (PPS) is one of the principal systems used in nuclear facilities for protection against terrorist attacks and the possibility of bombing, sabotage, and theft. This paper introduces the design of an advanced security system for nuclear waste warehouses against stealth terrorist attacks. This research paper takes a more original approach in addressing how a nuclear facility can detect and respond to a security incident. It does so through an interconnected security system that simultaneously practical and distinctive way. This PPS design allows the security system to achieve its main goals of detection and delay functions during any incident. Additionally, it places the intruder under attack from a successive defensive system which grants the response forces additional time to arrive and neutralize the threat completely.The proposed design of the security system will include dual-side infrared rays, high-intensive sound barriers, an intensive flashlight barrier, an electric chock system, and a smart camera surveillance system. Three consecutive processes, detection, deterrence, and defense will be operated around the waste storage building. This idea can be implemented by establishing an early detection system to detect any intrusion by terrorist attackers through a protected area surrounding the storage facility. The next step is deterring the attackers from completing their mission by placing an advanced deterrence system and putting delay elements along the path of their attack on the waste storage to complicate the mission. Then the defense system is operating to prevent terrorist progress. Finally, an evaluation process of the overall security system which has been designed by using EASI program is performed to verify the accepted level of physical protection, keep the security system component valid and updated, and follow procedures, rules and requirements of the nuclear regulatory authority at the national level and meet IAEA concept and recommendations at the international level.

An optimized deep belief network based pothole detection model for asphalt road

The poor quality of asphalt roads has a significant impact on driver safety, damages the mechanical structure of vehicles, increases fuel consumption, annoys passengers and is sometimes also responsible for accidents. Further, the poor quality of the road can be described as a rough surface and the presence of potholes. The potholes can be one of the main reasons for accident cause, increased fuel consumption and annoying passengers. Furthermore, the potholes can be of varied size, radiance effect, shadow and scales. Hence, the detection of potholes in asphalt roads can be considered a complex task and one of the serious issues regarding the maintenance of asphalt roads. This work focuses on the detection of the potholes in the asphalt roads. So in this work, a pothole detection model is proposed for accurate detection of potholes in the asphalt roads. The effectiveness of the proposed pothole detection model is tested over a set of real-world image datasets. In this study, the asphalt roads of the Delhi-NCR region are chosen and real-world images of these roads are collected through the smart camera. The final road image dataset consists of a total of 1150 images including 860 pothole images and the rest of are without pothole images. Further, the deep belief network is integrated into a proposed model for the detection of pothole images as a classification task and classified the images as pothole detected and not pothole. The experimental results of the proposed detection model are evaluated using accuracy, precision, recall, F1-Score and AUC parameters. These results are also compared with ANN, SVM, VGG16, VGG19 and InceptionV3 techniques. The simulation results showed that the proposed detection model achieves a 93.04% accuracy rate, 94.30% recall rate, 96.31% precision rate and 96.92% F1-Score rate than other techniques.

Optimum frequency tracking of the ultrasonic motor based on a three-level constant voltage drive

Abstract Ultrasonic motors are widely used in aerospace, smart cameras, and other applications. For ultrasonic motors to have the highest operating efficiency, the operating frequency of the motor needs to be adjusted in real time so that the motor’s operating frequency tracks the optimum frequency. Since the current research is limited to the optimum frequency tracking under constant velocity, the three-level drive parameters need to be adjusted to stabilize the velocity, which lacks precision and does not consider the frequency tracking under variable velocity. This paper explores an optimum frequency tracking scheme for ultrasonic motors under the three-level constant voltage input drive and proposes a variable step-size optimum frequency tracking scheme. The experimental results show that the optimum frequency of ultrasonic motors in variable velocity motion exists, and the proposed tracking scheme with three-level drive parameters can effectively improve the efficiency of the motor and significantly broaden the application scenario of the optimum frequency of ultrasonic motors.

Intelligent field monitoring system for cruciferous vegetable pests using yellow sticky trap images and an improved Cascade R-CNN

Cruciferous vegetables are important edible vegetable crops. However, they are susceptible to various pests during their growth process, which requires real-time and accurate monitoring of these pests for pest forecasting and scientific control. Hanging yellow sticky boards are a common way to monitor and trap those pests which are attracted to the yellow color. To achieve real-time, low-cost, intelligent monitoring of these vegetable pests on the boards, we established an intelligent monitoring system consisting of a smart camera, a web platform and a pest detection algorithm deployed on a server. After the operator sets the monitoring preset points and shooting time of the camera on the system platform, the camera in the field can automatically collect images of multiple yellow sticky boards at fixed places and times every day. The pests trapped on the yellow sticky boards in vegetable fields, Plutella xylostella, Phyllotreta striolata and flies, are very small and susceptible to deterioration and breakage, which increases the difficulty of model detection. To solve the problem of poor recognition due to the small size and breaking of the pest bodies, we propose an intelligent pest detection algorithm based on an improved Cascade R-CNN model for three important cruciferous crop pests. The algorithm uses an overlapping sliding window method, an improved Res2Net network as the backbone network, and a recursive feature pyramid network as the neck network. The results of field tests show that the algorithm achieves good detection results for the three target pests on the yellow sticky board images, with precision levels of 96.5, 92.2 and 75.0%, and recall levels of 96.6, 93.1 and 74.7%, respectively, and an F1 value of 0.880. Compared with other algorithms, our algorithm has a significant advantage in its ability to detect small target pests. To accurately obtain the data for the newly added pests each day, a two-stage pest matching algorithm was proposed. The algorithm performed well and achieved results that were highly consistent with manual counting, with a mean error of only 2.2%. This intelligent monitoring system realizes precision, good visualization, and intelligent vegetable pest monitoring, which is of great significance as it provides an effective pest prevention and control option for farmers.

Efficient Task-driven Video Data Privacy Protection for Smart Camera Surveillance System

As one of the most commonly used AIoT sensors, smart cameras and their supporting services, namely cloud video surveillance (CVS) systems, have brought great convenience to people’s lives. Recent CVS providers use different machine learning techniques to improve their services (regarded as tasks) based on the uploaded video. However, uploading data to the CVS providers may cause severe privacy issues. Existing works that remove privacy information could not achieve a high tradeoff between data usability and privacy, because the importance of information varies with the task. In addition, it is challenging to design a real-time privacy protection mechanism, especially in resource-constrained smart cameras. In this work, we design a task-driven and efficient video privacy protection mechanism for a better tradeoff between privacy and data usability. We use Class Activation Mapping to protect privacy while preserving data usability. To improve the efficiency, we utilize the motion vector and residual matrix produced during video codec. Our work outperforms the region of interest–based methods in data protection while preserving data usability. The attack accuracy drops 70%, while the task accuracy is comparable to those without protection (within ± 4%). The average protection frame rate of the High Definition video can exceed 16 fps+ even on a CPU.

TXAI-ADV: Trustworthy XAI for Defending AI Models against Adversarial Attacks in Realistic CIoT

Adversarial attacks are more prevalent in Consumer Internet of Things (CIoT) devices (i.e., smart home devices, cameras, actuators, sensors, and micro-controllers) because of their growing integration into daily activities, which brings attention to their possible shortcomings and usefulness. Keeping protection in the CIoT and countering emerging risks require constant updates and monitoring of these devices. Machine learning (ML), in combination with Explainable Artificial Intelligence (XAI), has become an essential component of the CIoT ecosystem due to its rapid advancement and impressive results across several application domains for attack detection, prevention, mitigation, and providing explanations of such decisions. These attacks exploit and steal sensitive data, disrupt the devices’ functionality, or gain unauthorized access to connected networks. This research generates a novel dataset by injecting adversarial attacks into the CICIoT2023 dataset. It presents an adversarial attack detection approach named TXAI-ADV that utilizes deep learning (Mutli-Layer Perceptron (MLP) and Deep Neural Network (DNN)) and machine learning classifiers (K-Nearest Neighbor (KNN), Support Vector Classifier (SVC), Gaussian Naive Bayes (GNB), ensemble voting, and Meta Classifier) to detect attacks and avert such situations rapidly in a CIoT. This study utilized Shapley Additive Explanations (SHAP) techniques, an XAI technique, to analyze the average impact of each class feature on the proposed models and select optimal features for the adversarial attacks dataset. The results revealed that, with a 96% accuracy rate, the proposed approach effectively detects adversarial attacks in a CIoT.

Real-Time Recognition of the User's Arm Gestures in 2D Space with a Smart Camera

Real-Time Recognition of the User's Arm Gestures in 2D Space with a Smart Camera