People Detection Research Articles

Bridging the Appearance Domain Gap in Elderly Posture Recognition with YOLOv9

Accurate posture detection of elderly people is crucial to improve monitoring and provide timely alerts in homes and elderly care facilities. Human posture recognition is experiencing a great leap in performance with the incorporation of deep neural networks (DNNs) such as YOLOv9. Unfortunately, DNNs require large amounts of annotated data for training, which can be addressed by using virtual reality images. This paper investigates how to address the appearance domain that lies between synthetic and natural images. Therefore, four experiments (VIRTUAL–VIRTUAL; HYBRID–VIRTUAL; VIRTUAL–REAL; and HYBRID–REAL) were designed to assess the feasibility of recognising the postures of virtual or real elderly people after training with virtual and real images of elderly people. The results show that YOLOv9 achieves the most outstanding accuracy of 98.41% in detecting and discriminating between standing, sitting, and lying postures after training on a large number of virtual images complemented by a much smaller number of real images when testing on real images.

Design of a Low-Cost AI System for the Modernization of Conventional Cars

Artificial intelligence techniques are beginning to be implemented in most areas. In the particular case of automobiles, new cars include integrated applications, such as cameras in different configurations, including in the rear of the car to provide assistance while reversing, as well as front and side cameras; these applications also include different configurations of sensors that provide information to the driver, such as objects approaching from different directions, such as from the front and sides. In this paper, we propose a practical and low-cost methodology to provide solutions using artificial intelligence techniques, as is the purpose of YOLO architecture, version 3, using hardware based on Nvidia’s Jetson TK1 architecture, and configurations in conventional cars. The results that we present demonstrate that these technologies can be applied in conventional cars, working with independent power to avoid causing problems in these cars, and we evaluate their application in the detection of people and cars in different situations, which allows information to be provided to the driver while performing maneuvers. The methodology that we provide can be replicated and scaled according to needs.

People Detection Using Artificial Intelligence with Panchromatic Satellite Images

The detection of people in urban environments from satellite imagery can be employed in a variety of applications, such as urban planning, business management, crisis management, military operations, and security. A WorldView-3 satellite image of Prague was processed. Several variants of feature-extracting networks, referred to as backbone networks, were tested alongside the Faster R–CNN model. This model combines region proposal networks with object detection, offering a balance between speed and accuracy that is well suited for dense and varied urban environments. Data augmentation was used to increase the robustness of the models, which contributed to the improvement of classification results. Achieving a high level of accuracy is an ongoing challenge due to the low spatial resolution of available imagery. An F1 score of 54% was achieved using data augmentation, a 15 cm buffer, and a maximum distance limit of 60 cm.

SMART CROWD MONITORING SYSTEM USING IOT-BASED YOLO-GHOST

Internet of Things (IoT) devices offer a smart, sophisticated, real-time surveillance solution for public spaces. However, due to instantaneous lighting changes and varying viewing angles, counting and tracking the people in crowded scenes is a challenging problem. To combat these issues, a novel YOLO-CROWD is proposed for a smart crowd-monitoring system using YOLO-GHOST. Initially, an Internet Protocol (IP) based camera was used to monitor and capture crowds in a video sequence. The captured video sequences are converted into frames and passed to the server via the Internet. The recorded frames are given to the YOLO-GHOST classification module to perform people detection and counting. Finally, the detected output results are transferred to the surveillance monitoring center's server. The YOLO-CROWD technique is simulated by using MATLAB. The effectiveness of the proposed YOLO-CROWD technique is assessed using evaluation metrics such as accuracy, precision, recall, sensitivity, F1-score, and mean average precision. The experimental results show that the accuracy of the YOLO-CROWD has increased to up to 99.95 %, proving that its intended use is for accurate crowd detection. The detection accuracy of the proposed method is 84.9 %, 87.58 %, 93.91 %, and 97.72 % better than existing EABeD, LCDnet, CDEM-M, and Public Vision, respectively.

Healthcare Professionals' Perspectives on Post-Diagnostic Care for People with Vascular Cognitive Impairment: When Help Is Needed in a "No-Man's Land".

Post-diagnostic care for people with vascular cognitive impairment (VCI) typically involves multiple professions and disjointed care pathways not specifically designed to aid VCI needs. Exploring perspectives of healthcare professionals on post-diagnostic care for people with VCI. We conducted a qualitative focus group study. We used purposive sampling to include healthcare professionals in different compositions of primary and secondary care professionals per focus group. Thematic saturation was reached after seven focus groups. Transcripts were iteratively coded and analyzed using inductive thematic analysis. Forty participants were included in seven focus groups (4-8 participants). Results showed knowledge and awareness of VCI as prerequisites for adequate post-diagnostic care, and for pre-diagnostic detection of people with VCI (theme 1). In light of perceived lack of differentiation between cognitive disorders, participants shared specific advice regarding post-diagnostic care for people with VCI and informal caregivers (theme 2). Participants thought current care for VCI was fragmented and recommended further integration of care and collaboration across settings (theme 3). People with VCI and their caregivers risk getting stuck in a "no man's land" between post-diagnostic care pathways; challenges lie in acknowledgement of VCI and associated symptoms, and alignment between healthcare professionals. Education about the symptoms and consequences of VCI, to healthcare professionals, people with VCI and caregivers, may increase awareness of VCI and thereby better target care. Specific attention for symptoms common in VCI could further tailor care and reduce caregiver burden. Integration could be enhanced by combining expertise of dementia and stroke/rehabilitation pathways.

INTEGRATED CONVOLUTIONAL NEURAL NETWORKS APPLICATION FOR ROAD TRAFFIC EFFICIENCY

Road traffic is a problem that every country has to face for the benefit of its population. Everything related to the transportation of people, goods, and services can be delayed due to a deplorable infrastructure. From minor tasks like individual shopping to important services like ambulances and fire trucks, all vehicles in Romania will eventually pass through a traffic light-controlled intersection that is not managed in real-time and that is also highly inefficient. At such a location, it is critical for traffic flow to be as smooth as possible to prevent time losses, and for traffic information to be stored into a cloud for process optimization. The purpose of the developed application is to reduce the real-time traffic jams to allow the optimal vehicles movement in safe conditions for all traffic participants. Using deep learning (DL) and artificial intelligence (AI) techniques, the application is based on two convolutional neural networks (CNN’s), respectively one for real-time detection of cars and people, and another one for real-time detection of cars license plates. The cars detections with CNN’s is important for road traffic optimization, for the efficient pedestrian detection necessary for crosswalks, and the car licence plates detection necessary for preventing traffic violations. The developed application contains several modules that operate based on deep learning and devices that communicate with each other, ensuring efficient data saving in cloud, data visualization, real-time traffic light control, and the information processing from traffic cameras.

Ten-year risk assessment for cardiovascular disease & associated factors among adult Indians (aged 40-69 yr): Insights from the National Noncommunicable Disease Monitoring Survey (NNMS).

Background & objectives Cardiovascular diseases (CVDs) are extremely prevalent in India, making early detection of people at high risk for CVDs and prevention crucial. This study aimed to estimate CVD risk distribution in older adults (40-69 yr) in India using WHO's non-laboratory risk chart and identify factors associated with elevated CVD risk (10%). Methods The current study used a nationally representative sample of 40-69 yr adults in India. The population's 10-yr CVD risk was defined as very low-to-low (10%), moderate (10-20%) and high to very high (>20%). We attempted univariable and multivariable logistic regressions to identify factors related to higher CVD risk (≥10%). Results Out of 4480 participants, 50 per cent were younger participants (40-49 years). The proportions of the population with very low to low, moderate and high to very high CVD risk were 84.9, 14.4 and 0.7 per cent, respectively. The estimated 10-year CVD risk was higher for people with unemployed [Adjusted Odds Ratio (AOR): 5.12; 95% Confidence Interval (CI): 3.63, 7.24], followed by raised blood glucose (AOR: 1.81; 95%CI: 1.39, 2.34). Interpretation & conclusions The non-laboratory-based chart proves valuable in low-resource settings, especially at the primary healthcare level, facilitating efficient CVD risk assessment and resource allocation. Further research is needed to explore the association of second-hand smoke with CVD risk in the Indian population.

Dense Pedestrian Detection Based on GR-YOLO.

In large public places such as railway stations and airports, dense pedestrian detection is important for safety and security. Deep learning methods provide relatively effective solutions but still face problems such as feature extraction difficulties, image multi-scale variations, and high leakage detection rates, which bring great challenges to the research in this field. In this paper, we propose an improved dense pedestrian detection algorithm GR-yolo based on Yolov8. GR-yolo introduces the repc3 module to optimize the backbone network, which enhances the ability of feature extraction, adopts the aggregation-distribution mechanism to reconstruct the yolov8 neck structure, fuses multi-level information, achieves a more efficient exchange of information, and enhances the detection ability of the model. Meanwhile, the Giou loss calculation is used to help GR-yolo converge better, improve the detection accuracy of the target position, and reduce missed detection. Experiments show that GR-yolo has improved detection performance over yolov8, with a 3.1% improvement in detection means accuracy on the wider people dataset, 7.2% on the crowd human dataset, and 11.7% on the people detection images dataset. Therefore, the proposed GR-yolo algorithm is suitable for dense, multi-scale, and scene-variable pedestrian detection, and the improvement also provides a new idea to solve dense pedestrian detection in real scenes.

Artificial intelligence-assisted adenoma detection in people with Lynch syndrome

Artificial intelligence-assisted adenoma detection in people with Lynch syndrome

Rapid CD4 cell count determination and Cryptococcus and Histoplasma antigen detection in people living with HIV: Implementation of a package of care strategy in a pilot study, Argentina

ObjectivesDespite advancements in HIV diagnosis and treatment, advanced HIV disease (AHD) is still a significant concern worldwide, especially in countries with a high percentage of undiagnosed cases and late-stage diagnoses. MethodsA prospective pilot study was conducted in Buenos Aires, Argentina to assess the feasibility of implementing a package for people living with HIV, integrating a point-of-care clusters of differentiation (CD4) test, followed by rapid Cryptococcus and Histoplasma antigen (Ag) detection. ResultsA total of 105 people living with HIV were enrolled, during June 2021 to October 2021. The VISITECT CD4 Advanced Disease Lateral Flow Assay (CD4-LFA) (Accubio) classified 98 (93%) patients with AHD. Compared with flow cytometry, the CD4-LFA performed with a high sensitivity (100%) but low specificity (19%) and limited accuracy (47%). In the 98 patients classified with AHD using the CD4-LFA, 16 tested positive for any of the rapid Ag used, including 12 patients positive for the Histoplasma Ag test and four positive for Cryptococcus Ag; all four patients with positive Cryptococcus Ag in sera and were diagnosed with meningitis. In the 30-day follow-up, one death was recorded. ConclusionsThe CD4-LFA correctly classified all patients with CD4 ≤200 cells/µL by flow cytometry, but a high frequency of patients misclassified with AHD was recorded. We also observed a high prevalence of opportunistic fungal infections, as previously observed in the hospital where this pilot study was conducted; however, in contrast with those previous reports, mortality was lower. The study underscores the importance of scaling up comprehensive care strategies and collaborating with governmental and non-governmental partners to enhance access to essential diagnostic tools and treatments for people living with HIV. Further research with larger sample sizes is needed to validate these findings.

Advanced deep learning for masked individual surveillance

During Covid-19 pandemic, face masks have become a ubiquitous protective measure. This poses new challenges for surveillance systems that heavily rely on facial recognition. To address this critical issue, we present a novel enhanced surveillance system that leverages deep learning techniques to tackle two crucial tasks simultaneously: anomaly detection of masked individuals' activities and masked face completion for accurate recognition. For anomaly detection, we employ a custom-designed deep neural network capable of processing real-time video streams. Finding a dataset of anomaly events of masked individuals is a big challenge for us. We handle this challenge using efficient techniques such as Dlib library and other image processing techniques. The network is trained on a diverse dataset encompassing normal and abnormal activities of masked individuals, enabling it to identify suspicious behaviors effectively. The surveillance cameras will exchange information, using a suitable network protocol, about detected anomalies and share relevant image data to aid in decision-making and choose the best images for further processing. In the context of masked face completion, we present a novel architecture called CCGAN network that is a combination of convolutional neural network (CNN) and conditioned generative adversarial network (CGAN) to generate the hidden parts of the face in a form that is accurate and close to the original face shape, as shown in our results. We conduct extensive experiments on publicly available datasets, demonstrating superior performance in both anomaly detection and masked face completion tasks. We have achieved 90% accuracy for anomaly detection of masked people.

A Proposed Model for Fainting People Detection Using Media Pipe Technology

Nowadays, falls put patients' safety and health in danger in hospitals and assisted living facilities, especially at night. The camera system used in this study is the suggested model; it records video footage and, in real time, analyzes patterns using Media Pipe algorithms to detect fainting or falls. An instant alert is generated and sent to a mobile app connected to the camera system if such an event is detected. The caretakers, nurses, or security officers entrusted with providing assistance can communicate with each other using the smartphone application. They follow the message to a specific location. Lastly, the system aims to improve the quality of care and support for the senior population by accelerating the response to instances of fainting and falling through the use of computer vision technologies and real-time notifications.

User Requirements and Perceptions of a Sensor System for Early Stress Detection in People With Dementia and People With Intellectual Disability: Qualitative Study.

Timely detection of stress in people with dementia and people with an intellectual disability (ID) may reduce the occurrence of challenging behavior. However, detecting stress is often challenging as many long-term care (LTC) residents with dementia and residents with ID have communication impairments, limiting their ability to express themselves. Wearables can help detect stress but are not always accepted by users and are uncomfortable to wear for longer periods. Integrating sensors into clothing may be a more acceptable approach for users in LTC. To develop a sensor system for early stress detection that is accepted by LTC residents with dementia and residents with ID, understanding their perceptions and requirements is essential. This study aimed to (1) identify user requirements for a garment-integrated sensor system (wearable) for early stress detection in people with dementia and people with ID, (2) explore the perceptions of the users toward the sensor system, and (3) investigate the implementation requirements in LTC settings. A qualitative design with 18 focus groups and 29 interviews was used. Focus groups and interviews were conducted per setting (dementia, ID) and target group (people with dementia, people with ID, family caregivers, health care professionals). The focus groups were conducted at 3 time points within a 6-month period, where each new focus group built on the findings of previous rounds. The data from each round were used to (further) develop the sensor system. A thematic analysis with an inductive approach was used to analyze the data. The study included 44 participants who expressed a positive attitude toward the idea of a garment-integrated sensor system but also identified some potential concerns. In addition to early stress detection, participants recognized other potential purposes or benefits of the sensor system, such as identifying triggers for challenging behavior, evaluating intervention effects, and diagnostic purposes. Participants emphasized the importance of meeting specific system requirements, such as washability and safety, and user requirements, such as customizability and usability, to increase user acceptance. Moreover, some participants were concerned the sensor system could contribute to the replacement of human contact by technology. Important factors for implementation included the cost of the sensor system, added value to resident and health care professionals, and education for all users. The idea of a garment-integrated sensor system for early stress detection in LTC for people with dementia and people with ID is perceived as positive and promising by stakeholders. To increase acceptability and implementation success, it is important to develop an easy-to-use, customizable wearable that has a clear and demonstrable added value for health care professionals and LTC residents. The next step involves pilot-testing the developed wearable with LTC residents with dementia and residents with ID in clinical practice.

Kinetics of SARS-CoV-2 Viral Load in Hospitalized Patients.

The rapid and accurate detection of infectious people is crucial in controlling outbreaks. The aim of this study was to evaluate the kinetics of the viral load expressed as Ct in COVID-19 hospitalized patients. Nasopharyngeal swab specimens were collected for RT-PCR testing. Forty-one subjects were recruited, of which 48.8% developed severe symptoms and 51.2% showed milder symptoms. The distribution of Ct values measured from the symptom onset showed that the kinetics of the viral load decreased with increasing time. A Ct of 25 (high viral load) was reached after a mean of 9.9 ± 4.8 days from the symptom onset, without a significant difference between patients with severe (10.9 ± 5.7 days) and milder (9.0 ± 3.9 days) symptoms. In 65.8% of cases, a high viral load was maintained for more than 7 days from the symptom onset, especially in patients with severe symptoms (70.6%). A Ct of 30 (moderate viral load) and of 38 (low viral load) were reached after a mean of 16.1 ± 8.1 and 28.5 ± 22.4 days from the symptom onset, respectively, with a significant difference between patients with severe (Ct = 30:17.9 ± 9.8 days; Ct = 38:34.6 ± 29.6 days) and milder (Ct = 30:14.3 ± 5.8 days; Ct = 38:22.7 ± 9.9 days) symptoms. These results provide an understanding of the viral kinetics of SARS-CoV-2 and have implications for pandemic control strategies and practices.

Newly Diagnosed HIV Infection in Patients with Progressive Multifocal Leukoencephalopathy (Literature Review and Clinical Cases)

In recent years, in the Kyrgyz Republic, HIV infection is more often detected at an advanced stage based on of advanced opportunistic infections. Late detection of people living with HIV causes disability and death due to untimely initiation of antiretroviral therapy and late prevention of opportunistic infections. The most difficult secondary disease to diagnose in the advanced stage of HIV infection is progressive multifocal leukoencephalopathy, often with an unfavorable outcome. This article presents two newly identified clinical cases of HIV infection in patients with difficult-to-diagnose progressive multifocal leukoencephalopathy, which may increase doctors' alertness to these diseases.

Impact of trans-regional travel-related testing on epidemic spreading

The convenient transport network enables people to travel quickly and frequently across regions, but it also brings serious challenges to the emergency management of infectious diseases. This work extends the classical SEIR model by incorporating trans-regional population movements and contact infections during travel. It also includes cross-regional population testing and treatment measures to curb the spread of infectious diseases across regions. According to these characteristics, a Net-SEIHR model is established and its diffusion properties are deduced. Further, the cooperative allocation problem of limited detection resources and its control strategy are discussed. Numerical experiments show that the control measures can reduce the speed of trans-regional transmission of infectious diseases and the peak of treatment. In the case of limited testing resources, it is necessary to strengthen the detection of people in the inflow regions.

Design of interactive online teaching system for big data computer information system based on mobile sensor network

The combination of the Internet and computers has resulted in many new products, including smart homes. The emergence of smart homes can satisfy people's detection of the internal environment of the home, not only to ensure the continuity and safety of indoor monitoring, but also to provide convenience for people to manage their family life. This paper analyzes the current status of smart home research at home and abroad, and people's needs for homes, and builds and designs smart home systems based on Ethernet and ZigBee protocol wireless sensor networks connected to TCP/IP protocol. In this article, the ARM chip is used in the Android system to ensure that the system and functions are matched, and information is transmitted through the serial port of the development board. After receiving the corresponding sending signal, the wireless module will use ZigBee technology to send the corresponding signal to achieve the goal of controlling home appliances. We divides the control system into two parts, the Android Voice Control Platform and the data collection terminal. The former is based on the Android operating system and is designed for voice recognition control; the latter uses a ZigBee microcontroller to achieve control and feedback. This article uses the built-in voice control system of the smart home to subtly replace the conventional control method of the smart home. Moreover, the dialogue with users is very high, easy to use, low cost, and very practical. In addition, the research methods used in this article can also be applied to other related fields.

Question-answering system extracts information on injection drug use from clinical notes

BackgroundInjection drug use (IDU) can increase mortality and morbidity. Therefore, identifying IDU early and initiating harm reduction interventions can benefit individuals at risk. However, extracting IDU behaviors from patients’ electronic health records (EHR) is difficult because there is no other structured data available, such as International Classification of Disease (ICD) codes, and IDU is most often documented in unstructured free-text clinical notes. Although natural language processing can efficiently extract this information from unstructured data, there are no validated tools.MethodsTo address this gap in clinical information, we design a question-answering (QA) framework to extract information on IDU from clinical notes for use in clinical operations. Our framework involves two main steps: (1) generating a gold-standard QA dataset and (2) developing and testing the QA model. We use 2323 clinical notes of 1145 patients curated from the US Department of Veterans Affairs (VA) Corporate Data Warehouse to construct the gold-standard dataset for developing and evaluating the QA model. We also demonstrate the QA model’s ability to extract IDU-related information from temporally out-of-distribution data.ResultsHere, we show that for a strict match between gold-standard and predicted answers, the QA model achieves a 51.65% F1 score. For a relaxed match between the gold-standard and predicted answers, the QA model obtains a 78.03% F1 score, along with 85.38% Precision and 79.02% Recall scores. Moreover, the QA model demonstrates consistent performance when subjected to temporally out-of-distribution data.ConclusionsOur study introduces a QA framework designed to extract IDU information from clinical notes, aiming to enhance the accurate and efficient detection of people who inject drugs, extract relevant information, and ultimately facilitate informed patient care.

Adenoma detection rate as a quality indicator for colonoscopy.

Objective: To determine the quality of colonoscopy in terms of adenoma detection rate in the existing local settings. Study Design: Cross-sectional study. Setting: Department of Gastroenterology, Liaquat National Hospital, Karachi, Pakistan. Period: February 2022 to August 2022. Methods: We analyzed patients having age 18 years or older, of any gender, and who had colonoscopy indications and were required to have a screening colonoscopy for diagnostic purposes. The number of procedures with at least one polyp or adenoma was classified as the polyp detection rate (PDR) and adenoma detection rate (ADR). Result: Total 210 patients were enrolled into the study with median age of 41.5 (29.8-56.3) years and majority were males (n=46, 69.5%). The most frequent indication for colonoscopy was constipation (n=72, 34.3%) followed by bleeding (n=78, 37.1%), chronic diarrhea (n=36, 17.1%), both constipation and diarrhea (n=21, 10%) and altered bowel habits (n=3, 1.4%). PDR and ADR were 16.2% (n=34) and 10% (n=21) respectively. Age was significantly higher among patients who were found to have polyp (p<0.001) and adenomas (p<0.001). Frequency of polyps (p=0.010) and adenomas (p=0.007) were significantly higher in male patients. Conclusion: The current research discovered a low rate of adenoma detection in people receiving colorectal cancer screening, and it did not meet the quality benchmarks that have been associated with higher rates of early diagnosis and cancer prevention.

Development of artificial intelligence edge computing based wearable device for fall detection and prevention of elderly people

Elderly falls are a major concerning threat resulting in over 1.5–2 million elderly people experiencing severe injuries and 1 million deaths yearly. Falls experienced by Elderly people may lead to a long‐term negative impact on their physical and psychological health conditions. Major healthcare research had focused on this lately to detect and prevent the fall. In this work, an Artificial Intelligence (AI) edge computing based wearable device is designed and developed for detection and prevention of fall of elderly people. Further, the various deep learning algorithms such as Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU) are utilized for activity recognition of elderly. Also, the CNN-LSTM, RNN-LSTM and GRU-LSTM with and without attention layer respectively are utilized and the performance metrics are analyzed to find the best deep learning model. Furthermore, the three different hardware boards such as Jetson Nano developer board, Raspberry PI 3 and 4 are utilized as an AI edge computing device and the best deep learning model is implemented and the computation time is evaluated. Results demonstrate that the CNN-LSTM with attention layer exhibits the accuracy, recall, precision and F1_Score of 97%, 98%, 98% and 0.98 respectively which is better when compared to other deep learning models. Also, the computation time of NVIDIA Jetson Nano is less when compared to other edge computing devices. This work appears to be of high societal relevance since the proposed wearable device can be used to monitor the activity of elderly and prevents the elderly falls which improve the quality of life of elderly people.