Browse Items (11809 total)
Sort by:
-
Recent Advances in Analytical Techniques for Antidepressants Determination in Complex Biological Matrices: A Review
Depression is one of the most prevalent but severe of mental disorders, affecting thousands of individuals across the globe. Depression, in its most extreme form, may result in self-harm and an increased likelihood of suicide. Antidepressant drugs are first-line medications to treat mental disorders. Unfortunately, these medications are also prescribed for other in- and off-label conditions, such as deficit/hyperactivity disorders, attention disorders, migraine, smoking cessation, eating disorders, fibromyalgia, pain, and insomnia. This results in an increase in the use of antidepressant medications, leading to clinical and forensic overdose cases that could be either accidental or deliberate. The findings revealed that people who used antidepressants had a 33% greater chance of dying sooner than expected, compared to those who did not take the medications. Analytical techniques for precisely identifying and detecting antidepressants and their metabolic products in a variety of biological matrices are greatly needed to be developed and made available. Hence, this study attempts to discuss various analytical techniques used to identify and determine antidepressants in various biological matrices, which include urine, blood, oral fluid (saliva), and tissues, which are commonly encountered in clinical and forensic science laboratories. The Author(s) 2023. -
Recall of advertisements placed during Indian crime shows /
Brad J. Bushman conducted a study in 1998 that studied the effects of watching televised violence on memory. Bushman conducted experiments to find out if the viewers remembered the advertisements after watching violence on television. Results showed that violence had a negative effect on recall of the brand and the advertisement message. The present research sought to study the effect of violence in Indian crime shows on the recall of the brand, the product category and the tag lines used in the advertisements. -
Rebuilding the Capabilities for Post COVID-19 Pandemic: Issues and Challenges of Bangalore Model of Development
The pace of urbanization has achieved considerable momentum in recent years with 34.93 per cent of India's population living in urban areas. However, the COVID - 19 pandemic has severely affected urban development with adverse effects on people's mobility, consumption level, health and poverty. Bangalore, the capital of Karnataka and the third largest city in India, has a population of 11 million and contributes more than one third of the state's GDP. The expansion of certain sectors including Information Technology, infrastructure and spread of educational institutions has fueled Bangalore's rapid growth in the past three decades which has made it a regional superpower in India, if not South Asia. This paper explores the unique features of the 'Bangalore Model of Development' as a regional development model and provides a systematic introspection of its capabilities. It discusses the impact of the pandemic on the key driving forces of Bangalore Model and assesses the current government measures. The situation analysis with the policy prescriptions would help to strengthen and sustain the urban system during the postpandemic times. 2022 IEEE. -
-
-
Realization of Humanoid Doctor and Real-Time Diagnostics of Disease Using Internet of Things, Edge Impulse Platform, and ChatGPT
Humanoid doctor is an AI-based robot that featured remote bi-directional communication and is embedded with disruptive technologies. Accurate and real-time responses are the main characteristics of a humanoid doctor which diagnoses disease in a patient. The patient details are obtained by Internet of Things devices, edge devices, and text formats. The inputs from the patient are processed by the humanoid doctor, and it provides its opinion to the patient. The historical patient data are trained using cloud artificial intelligence platform and the model is tested against the patient sample data acquired using medical IoT and edge devices. Disease is identified at three different stages and analyzed. The humanoid doctor is expected to identify the diseases well in comparison with human healthcare professionals. The humanoid doctor is under-trusted because of the lack of a multi-featured accurate model, accessibility, availability, and standardization. In this letter, patient input, artificial intelligence, and response zones are encapsulated and the humanoid doctor is realized. The Author(s) under exclusive licence to Biomedical Engineering Society 2023. -
Realization of Green 5G Cellular Network Role in Medical Applications: Use of ChatGPT-AI
Wireless communication in medical applications improves patient monitoring, care coordination, early disease detection, and patient empowerment. It improves healthcare and patient outcomes. The design and configuration of a solar-powered emergency battery backup system for 5G telecommunication base stations, including medical applications, may vary depending on local climate, power requirements, and resources. In this connection, uninterrupted power supply to the base stations become crucial. The author utilizes the ChatGPT-AI features and prepared this comprehensive letter for realizing the role of sustainable practices towards climatic changes. 2023, The Author(s) under exclusive licence to Biomedical Engineering Society. -
Real-World Application of Machine Learning and Deep Learning
The world today is running on the latest computer technologies and one of those is machine learning. The real life example that most of us know is speech recognition. Google Assistant is the common example for this Speech recognition. This google assistant is not only limited till 'Ok Google', but it responds to all your questions in a smart way. It can manage all your calls or can book appointments. Imagine you fell down while de-boarding a bus. So, Next time you take care so that you don't fall that is something that your brain has interpreted from your past experience. This is what exactly deep learning is, it imitates human brain works. Deep learning is sub-branch of machine learning. It is able to build all new things based on its previous experiences. Many of us have heard about driverless cars and medical diagnosis. Recently google has developed a new technology where all your cardiovascular events can be predicted by eye scan so, that doctors can get a clear view of what is inside the body of a patient. These all are developed using machine learning. It has a capability to change the human world into a complete robotic world. Anyways, it also has its own disadvantages. This article discusses about those, Scope of machine learning, its Market potential, financial growth and Current applications of machine learning. 2019 IEEE. -
Real-time video segmentation using a vague adaptive threshold
For the last two decades, video shot segmentation has been a widely researched topic in the field of content-based video analysis (CBVA). However, over the course of time, researchers have aimed to improve upon the existing methods of shot segmentation in order to gain accuracy. Video shot segmentation or shot boundary analysis is a basic and vital step in CBVA, since any error incurred in this step reduces the precision of the other steps. The shot segmentation problem assumes greater proportions when detection is preferred in real time. A spatiotemporal fuzzy hostility index (STFHI) is proposed in this work which is used for edge detection of objects occurring in the frames of a video. The edges present in the frames are treated as features. Correlation between these edge-detected frames is used as a similarity measure. In a real-time scenario, the incoming images are processed and the similarities are computed for successive frames of the video. These values are assumed to be normally distributed. The gradients of these correlation values are taken to be members of a vague set. In order to obtain a threshold after defuzzification, the true and false memberships of the elements are computed using a novel approach. The threshold is updated as new frames are buffered in and is referred to as the vague adaptive threshold (VAT). The shot boundaries are then detected based on the VAT. The VAT for detecting the shot boundaries is determined by using the three-sigma rule on the defuzzified membership values. The effectiveness of the real-time video segmentation method is established by an experimental evaluation on a heterogeneous test set, comprising videos with diverse characteristics. The test set consists of videos from sports, movie songs, music albums, and documentaries. The proposed method is seen to achieve an average F1 score of 0.992 over the test set consisting of 15 videos. Videos from the benchmark TRECVID 2001 are selected for comparison with other state-of-the-art-methods. The proposed method achieves very high precision and recall, with an average F1 score of 0.939 on the videos chosen from the TRECVID 2001 dataset. This is a substantial improvement over the other existing methods. 2020 Elsevier Inc. -
Real-Time Traffic Sign Detection Under Foggy Condition
Traffic congestion becomes high in urban areas and using public and private transportation services. The image of traffic signs gets affected by fog, and the detection of traffic signs has become difficult. To solve this issue, the machine learning technique has been used. Convolution neural network helps to solve real-time problems; hence, it can be used in the study for detecting traffic signs under foggy condition. The study results revealed that the model network has accuracy of 99.8%, and the proposed algorithm detects a traffic sign under foggy conditions in 2s per frame. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Real-time Traffic Prediction in 5G Networks Using LSTM Networks
This research explores the application of Long Short-Term Memory (LSTM) networks for real-time traffic prediction within 5G networks, aiming to address the critical need for accurate prediction models in dynamic network environments. Leveraging the sequential learning capabilities of LSTM networks, the proposed methodology encompasses dataset preparation, model architecture design, training, and evaluation. Experimental results demonstrate the effectiveness of the LSTM-based prediction model in capturing temporal dependencies and providing reliable predictions across various prediction horizons. While promising, further research is warranted to enhance the model's performance and address remaining challenges. This study contributes to advancing the state-of-the-art in traffic prediction methodologies, facilitating more efficient network management and optimization in 5G environments. 2024 IEEE. -
Real-Time State of Charge Prediction Model for Electric Two-Wheeler
To maximise the efficiency and performance of electric vehicles, traction battery State of Charge (SoC) must be accurately predicted. In this work, a prediction model for traction battery State of Charge estimation is developed in real time. The traction battery powers an electric two-wheeler through a predetermined drive cycle. To produce accurate state-of-charge forecasts, the predictive model considers several input characteristics, such as temperature, voltage, and current. This research is crucial for fostering effective energy management and improving the safety and dependability of electric two-wheelers. Open-circuit voltage (OCV) and coulomb counting are two commonly utilised techniques used to evaluate the state of charge prediction model. These techniques act as standards for assessing the developed Neural Network model prediction, the model's dependability and accuracy. The model's usefulness and its potential to outperform the current State of Charge estimating techniques are demonstrated by comparing the state-of-charge predictions from the model with these standard methods. 2024 IEEE. -
Real-time Litter Recognition Using Improved YOLOv4 Tiny Algorithm
Littered roads have become a familiar sight in India. The main reason is the increasing population and inefficient waste disposal system. Since garbage collectors cannot pick litter in all the places, there is a need for an efficient way to detect it. Hence, a machine learning-based object detection model is used. In this, we have applied an improved YOLOv4-Tiny algorithm to detect the garbage, classify it and make the detection process easier on custom datasets. We have improved the algorithm in terms of the object prediction time, this is done by replacing a max pooling layer with one of two layers present in a fully connected layer. When an input is given, the algorithm detects the litter in the image with a bounding box around it along with the label and confidence score. The proposed model reduces the prediction time by 0.517 milliseconds less than the original algorithm employed which concludes that the object is predicted faster. 2022 IEEE. -
Real-time human action prediction using pose estimation with attention-based LSTM network
Human action prediction in a live-streaming videos is a popular task in computer vision and pattern recognition. This attempts to identify activities in an image or video performed by a human. Artificial intelligence(AI)-based technologies are now required for the security and human behaviour analysis. Intricate motion patterns are involved in these actions. For the visual representation of video frames, conventional action identification approaches mostly rely on pre-trained weights of various AI architectures. This paper proposes a deep neural network called Attention-based long short-term memory (LSTM) network for skeletal based activity prediction from a video. The proposed model has been evaluated on the BerkeleyMHAD dataset having 11 action classes. Our experimental results are compared against the performance of the LSTM and Attention-based LSTM network for 6 action classes such as Jumping, Clapping, Stand-up, Sit-down, Waving one hand (Right) and Waving two hands. Also, the proposed method has been tested in a real-time environment unaffected by the pose, camera facing, and apparel. The proposed system has attained an accuracy of 95.94% on BerkeleyMHAD dataset. Hence, the proposed method is useful in an intelligent vision computing system for automatically identifying human activity in unpremeditated behaviour. The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024. -
Real-Time Fire Detection Through the Analysis of Surveillance Videos
The Forest Fire Detection System is an intelligent system that can detect forest fires and alert authorities in real-time. It uses a YOLOv5 deep learning algorithm to process live video feeds captured by a web camera which is trained with the sizable dataset of inputs to locate the fire accurately, making it an ideal choice for real-time fire detection in the forest. Upon detecting a fire, the system sends an email alert to a designated email address, containing a picture of the fire and location information. The email alert system is built using the standard SMTP protocol, which ensures that the message is delivered to the recipient in a timely and reliable manner. The system is also equipped with a speaker that triggers an alarm upon detecting a fire. The alarm is designed to alert people in the vicinity of the fire so that they can take the necessary action. It is activated using the Pygame library, a collection of Python modules specifically crafted for game development across multiple platforms. Overall, the Forest Fire Detection System is a fast, efficient, and accurate system that can help prevent the spread of forest fires. It is an intelligent system that can detect fires quickly and send alerts to authorities, giving them the information they need to take the necessary action to control the fire. The system is built using a web camera, a computer, and a speaker, making it easy to install and maintain. 2024 IEEE. -
Real-Time Cyber-Physical Risk Management Leveraging Advanced Security Technologies
Conducting an in-depth study on algorithms addressing the interaction problem in the fields of machine learning and IoT security involves a meticulous evaluation of performance measures to ensure global reliability. The study examines key metrics such as accuracy, precision, recall, and F1 scores across ten scenarios. The highly competitive algorithms showcase accuracy rates ranging from 95.5 to 98.2%, demonstrating their ability to perform accurately in various situations. Precision and recall measurements yield similar information about the model's capabilities. The achieved balance between accuracy and recovery, as determined by the F1 tests ranging from 95.2 to 98.0%, emphasizes the practical importance of data transfer in the proposed method. Numerical evaluation, in addition to an analysis of overall performance metrics, provides a comprehensive understanding of the algorithm's performance and identifies potential areas for improvement. This research leads to advancements in the theoretical vision of machine learning for IoT protection. It offers real-world insights into the practical use of robust models in dynamically changing situations. As the Internet of Things environment continues to evolve, the study's results serve as crucial guides, laying the foundation for developing strong and effective security systems in the realm of interaction between virtual and material reality. The Author(s) 2024. -
Real-time architectural efforts in building a social network using NOSQL databases
Relational database management systems (RDBMS) today are the predominant technology for storing structured data in web and business applications. Along with the increasing size of the datasets, the number of accesses and operations performed increases. This growth, enhanced by the proliferation of social networks, led to a depletion of traditional relational databases that were commonly used to solve a wide range of problems. -
Real-Time Application with Data Mining and Machine Learning
Data mining and machine learning are the most expressive research and application domain. All real-time application directly or indirectly depends on data mining and machine learning. There are manyrelevantfields, like data analysis in finance,retail, telecommunications sector, analyzing biological data, otherscientific uses, and intrusiondetection.The most expressive research and application domain is data mining and machine learning. Data mining and machine learning are used in all real-time applications, whether directly or indirectly. Data analysis in finance, retail, telecommunications, biological data analysis, extra scientific applications, and intrusion detection are just a few exampleswhere it can be used. Because it captures a lot of data from sales, client purchase histories, product transportation, consumption, and services, DM has a lot of applications in the retail industry. It's only logical that the amount of data collected will continue to climb as the Internet's accessibility, cost, and popularity increase. In the retail industry, DM assists in the detection of customer buying behaviors and trends, resulting in improved customer service and increased customer retention and satisfaction. 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG. -
Real-Time Application of Document Classification Based on Machine Learning
This research has been performed, keeping a real-time application of document (multi-page, varying length, scanned image-based) classification in mind. History of property title is captured in various documents, recorded against the said property in all the countries across the world. Information of the property, starting from ownership to the conveyance, mortgage, refinance etc. are buried under these documents. This is by far a human driven process to manage these digitized documents. Categorization of the documents is the primary step to automate the management of these documents and intelligent retrieval of information without or minimal human intervention. In this research, we have examined a popular, supervised machine learning technique called, SVM (support vector machine) with a heterogeneous data set of six categories of documents related to property. The model obtained an accuracy of 88.06% in classifying over 988 test documents. 2020, Springer Nature Switzerland AG.