A Novel Algorithm YOLOv4-Mini to Detection Automobile Rim Weld

Agricultural factories utilize advanced facilities and technologies to cultivate crops in a controlled environment, enhancing operational yields and reducing reliance on natural resources. This is crucial for ensuring a stable supply of agricultural products year-round and plays a significant role in the transformation of agricultural modernization. Automated Guided Vehicles (AGVs) are commonly employed in agricultural factories due to their low ownership costs and high efficiency. However, small embedded devices on AGVs face significant challenges in managing multiple tasks while maintaining the required timeliness. Multi-task learning (MTL) is increasingly employed to enhance the efficiency and performance of detection models in joint detection tasks, such as lane-line detection, pedestrian detection, and obstacle detection. The YOLOP (You Only Look for Panoptic Driving Perception) model demonstrates strong performance in simultaneously addressing these tasks; detecting lane lines in changeable agricultural factory scenarios is yet a challenging task, limiting the subsequent accurate planning and control of AGVs. This paper proposes a feedback-based network for joint detection tasks (MTNet) that simultaneously detects pedestrians, automated guided vehicles (AGVs), and QR codes, while also performing lane-line segmentation. This approach addresses the challenge faced by using embedded devices mounted on AGVs, which are unable to run multiple models for different tasks in parallel due to limited computational resources. For lane-line detection tasks, we also propose an improved YOLOP lane-line detection algorithm based on feature shift aggregation. Homemade datasets were used for training and testing. Comparative experiments of our model with different models in the target-detection and lane-line detection tasks, respectively, show the progressiveness of our model. Surprisingly, we also obtained a significant improvement in the model’s processing speed. Furthermore, we conducted ablation experiments to assess the effectiveness of our improvements in lane-line detection, all of which outperformed the original detection model.

The Internet of Things (IoT) is revolutionizing healthcare by integrating smart devices and sensors into medical practice. These devices can monitor patient health remotely, collect real-time data, and enable healthcare providers to deliver more personalized and timely care. From wearable fitness trackers that monitor vital signs to smart pills that track medication adherence, IoT in healthcare enhances patient outcomes, improves healthcare delivery efficiency, and reduces costs. An electrocardiogram (ECG) is an essential diagnostic tool for determining heart -related problems. In this research, we have developed a real -time ECG data acquisition method using small devices like Raspberry Pi Pico (RPiPico) and AD8232 ECG sensors. Our study analyses the ECG signal transmission on small embedded devices and a standalone machine so that when such critical data is transmitted on cloud resources via edge devices, the delay taken to transmit the data impacts the significant decisions of embedding the algorithms in resource-constrained devices. We analyzed the delays while the data was transmitted through the MQTT broker. We have also compared CPU and memory usage running on small and standalone devices. We observe that when the broker runs on the Rpi, it gives less delay, CPU utilization, and memory usage compared to the laptop as a broker. There is no packet loss in both cases. Thus, transferring critical data via a small processor like RPi with an MQTT broker is faster than transferring it through a general laptop processor with an MQTT broker.

Distracted driving is a critical safety issue that leads to numerous fatalities and injuries worldwide. This study addresses the urgent need for efficient and real-time machine learning models to detect distracted driving behaviors. Leveraging the Pretrained-YOLOv8 (P-YOLOv8) model, a real-time object detection system is introduced, optimized for both speed and accuracy. This approach addresses the computational constraints and latency limitations commonly associated with conventional detection models. The study demonstrates P-YOLOv8's versa-tility in both object detection and image classification tasks using the Distracted Driver Detection dataset from state farm, which includes 22,424 images across ten behavior categories. Our research explores the application of P-YOLOv8 for image classification, evaluating its performance compared to deep learning models such as VGG16, VGG19, and ResNet. Some traditional models often struggle with low accuracy, while others achieve high accuracy but come with high computational costs and slow detection speeds, making them unsuitable for real-time applications. P-YOLOv8 addresses these issues by achieving competitive accuracy with significant computational cost and efficiency advantages. In particular, P-YOLOv8 generates a lightweight model with a size of only 2.84 MB and a lower number of parameters, totaling 1, 451, 098, due to its innovative architecture. It achieves a high accuracy of 99.46% with this small model size, opening new directions for deployment on inexpensive and small embedded devices using Tiny Machine Learning (TinyML). The experimental results show robust performance, making P-YOLOv8 a cost-effective solution for real-time deployment. This study provides a detailed analysis of P-YOLOv8's architecture, training, and performance benchmarks, highlighting its potential for real-time use in detecting distracted driving.

With the wide adoption of AI applications, there is a pressing need of\nenabling real-time neural network (NN) inference on small embedded devices, but\ndeploying NNs and achieving high performance of NN inference on these small\ndevices is challenging due to their extremely weak capabilities. Although NN\npartitioning and offloading can contribute to such deployment, they are\nincapable of minimizing the local costs at embedded devices. Instead, we\nsuggest to address this challenge via agile NN offloading, which migrates the\nrequired computations in NN offloading from online inference to offline\nlearning. In this paper, we present AgileNN, a new NN offloading technique that\nachieves real-time NN inference on weak embedded devices by leveraging\neXplainable AI techniques, so as to explicitly enforce feature sparsity during\nthe training phase and minimize the online computation and communication costs.\nExperiment results show that AgileNN's inference latency is >6x lower than the\nexisting schemes, ensuring that sensory data on embedded devices can be timely\nconsumed. It also reduces the local device's resource consumption by >8x,\nwithout impairing the inference accuracy.\n

<span lang="EN-US">Smart healthcare is a program that provides services to patients and healthcare assistants at any time, it is used smartly with cloud computing. A smart mobile device can collect data from a wide range of receiving sets (Wi-Fi and cloud), underwired sensors, the internet of things (IoT). It permits doctors to monitor and diagnose patients with small embedded devices that are unobtrusive, wireless, lightweight, and capable of detecting, interpreting and communi-cating physiological signs easily in a timely manner without the need for a doctor present. This paper develops integrated smart records and reports and is shared via the cloud. A variety of small devices have been used that detect the object to be sensed and then tell the person monitoring the patient's condition, for example, body and room temperature, oxygen percentage, heart rate). This data is sent to the server via the Internet. To store details and monitor the patient through it. In the event that any error occurs in any of this data, such as a decrease or increase in the normal value, a signal will be sent via Telegram to the phone of the individual monitoring the patient's condition to alert and to take necessary action.</span>

Data clustering is crucial when it comes to data processing and analytics. The new clustering method overcomes the challenge of evaluating and extracting data from big data. Numerical or categorical data can be grouped. Existing clustering methods favor numerical data clustering and ignore categorical data clustering. Until recently, the only way to cluster categorical data was to convert it to a numeric representation and then cluster it using current numeric clustering methods. However, these algorithms could not use the concept of categorical data for clustering. Following that, suggestions for expanding traditional categorical data processing methods were made. In addition to expansions, several new clustering methods and extensions have been proposed in recent years. ROCK is an adaptable and straightforward algorithm for calculating the similarity between data sets to cluster them. This paper aims to modify the algorithm by creating a parameterized version that takes specific algorithm parameters as input and outputs satisfactory cluster structures. The parameterized ROCK algorithm is the name given to the modified algorithm (P-ROCK). The proposed modification makes the original algorithm more flexible by using user-defined parameters. A detailed hypothesis was developed later validated with experimental results on real-world datasets using our proposed P-ROCK algorithm. A comparison with the original ROCK algorithm is also provided. Experiment results show that the proposed algorithm is on par with the original ROCK algorithm with an accuracy of 97.9%. The proposed P-ROCK algorithm has improved the runtime and is more flexible and scalable.

Small embedded devices such as sensors and actuators will become the cornerstone of the Future Internet. To this end, generic, open and secure communication and service platforms are needed in order to be able to exploit the new business opportunities these devices bring. In this paper, we evaluate the current efforts to integrate sensors and actuators into the Internet and identify the limitations at the level of cooperation of these Internet-connected objects and the possible intelligence at the end points. As a solution, we propose the concept of Managed Ecosystem of Networked Objects, which aims to create a smart network architecture for groups of Internet-connected objects by combining network virtualization and clean-slate end-to-end protocol design. The concept maps to many real-life scenarios and should empower application developers to use sensor data in an easy and natural way. At the same time, the concept introduces many new challenging research problems, but their realization could offer a meaningful contribution to the realization of the Internet of Things.

As we race toward the Internet of Things (IoT), small embedded devices are increasingly becoming network-enabled. Often, these devices can't meet the computational requirements of current intrusion prevention mechanisms or designers prioritize additional features and services over security; as a result, many IoT devices are vulnerable to attack. We have developed an ultra-lightweight deep packet anomaly detection approach that is feasible to run on resource constrained IoT devices yet provides good discrimination between normal and abnormal payloads. Feature selection uses efficient bit-pattern matching, requiring only a bitwise AND operation followed by a conditional counter increment. The discrimination function is implemented as a lookup-table, allowing both fast evaluation and flexible feature space representation. Due to its simplicity, the approach can be efficiently implemented in either hardware or software and can be deployed in network appliances, interfaces, or in the protocol stack of a device. We demonstrate near perfect payload discrimination for data captured from off the shelf IoT devices.

There are many kinds of control networks, which have been used in various non-IP network areas, such as BA (building automation), FA (factory automation) and PA (process automation). They are introducing IP and face the issues of security and configuration complexity. The authors have proposed a model which intends to solve the issues while satisfying restrictions, i.e. small embedded devices, isolated networks and private naming system/name space, which are required when introducing new functionality into existing control networks. Secure bootstrap sequence and device-to-device communication using the chain of trust are the points of the model. This paper shows the practicability of the model through implementing the model experimentally.

The Internet of Things (IoT) paradigm aims to bring continuous interaction between small embedded devices and humans. The IoT has the potential to affect our daily lives and bring many benefits to society. Low-Power Wide-Area Networks (LPWAN) is a new IoT technology that offers long distance connectivity for a massive number of devices. LPWAN is a promising solution to enable complex IoT scenarios, such as smart cities and smart healthcare. LoRa is currently one of the leading LPWAN solutions available for public use. Due to the great number of connected devices and, in some cases, sensitive data transmitted in IoT networks, security is one of the main concerns in LPWAN. In this paper, we focus on the issues of key management in LoRaWAN. We propose a secure architecture for key management based on private blockchain and smart contracts in order to increase the levels of security and availability in LoRaWAN environments. In order to show the feasibility of the proposed architecture, a working prototype was implemented using open-source tools and commodity hardware.

This paper presents the first RISC-V vector processor design aimed at microcontrollers that uses the new RISC-V ‘V’ extension for vectors, part of the open-source RISC-V instruction set architecture (ISA). Being aimed at small embedded devices, it demonstrates a simpler method of parallel execution than traditional vector architectures to minimise logic. It has been synthesised for testing on an FPGA at 50MHz. Typical vector-compatible applications have been used as benchmarks. Performance has been improved by up to 5.8x for the demonstrated applications relative to a comparable scalar RISC-V processor, for an increase in FPGA resource utilisation of at most 2.6x.

MCA-net: A multi-task channel attention network for Myocardial infarction detection and location using 12-lead ECGs

Realtime and intelligent video surveillance via camera networks involve computation-intensive vision detection tasks with massive video data, which is crucial for safety in the edge-enabled industrial Internet of Things (IIoT). Multiple video streams compete for limited communication resources on the link between edge devices and camera networks, resulting in considerable communication congestion. It postpones the completion time and degrades the accuracy of vision detection tasks. Thus, achieving high accuracy of vision detection tasks under the communication constraints and vision task deadline constraints is challenging. Previous works focus on single camera configuration to balance the tradeoff between accuracy and processing time of detection tasks by setting video quality parameters. In this paper, an adaptive camera network self-configuration method (CANS) of video surveillance is proposed to cope with multiple video streams of heterogeneous quality of service (QoS) demands for edge-enabled IIoT. Moreover, it adapts to video content and network dynamics. Specifically, the tradeoff between two key performance metrics, i.e., accuracy and latency, is formulated as an NP-hard optimization problem with latency constraints. A low-complexity algorithm is proposed to solve the optimization problem based on greedy searching. Simulation on real-world surveillance datasets demonstrates that the proposed CANS method achieves low end-to-end latency (13 ms on average) with high accuracy (92%) with network dynamics, which validates its effectiveness.

With significant development of Internet of medical things (IoMT) and cloud-fog-edge computing, medical industry is now involving medical big data to improve quality of service in patient care. Karyotyping refers classifying human chromosomes. However, performing karyotyping task generally requires domain expertise in cytogenetics, long-period experience for high accuracy, and considerable manual efforts. An end-to-end chromosome karyotype analysis system is proposed over medical big data to automatically and accurately perform chromosome related tasks of detection, segmentation, and classification. Facing image data generated and collected by means of edge computing, we firstly utilize visual feature to generate chromosome candidates with Extremal Regions (ER) technology. Due to severe occlusion and cross overlapping, we utilize ring radius transform to cluster pixels with same property to approximate chromosome shapes. To solve the problem of unbalanced and small dataset by covering diverse data patterns, we proposed multidistributed generated advertising network (MD-GAN) to perform data enhancement by generating additional training samples. Afterwards, we fine-tune CNN for chromosome classification task by involving generated and sufficient training images. Through experiments in self-collected datasets, the proposed method achieves high accuracy in tasks of chromosome detection, segmentation, and classification. Moreover, experimental results prove that MD-GAN-based data enhancement contributes to classification results of CNN to a certain extent.

In substation lightning rod meter reading data taking, the classical object detection model is not suitable for deployment in substation monitoring hardware devices due to its large size, large number of parameters, and slow detection speed, while is difficult to balance detection accuracy and real-time requirements with the existing lightweight object detection model. To address this problem, this paper constructs a lightweight object detection algorithm, YOLOv5-Meter Reading Lighting (YOLOv5-MRL), based on the improved YOLOv5 model's speed while maintaining accuracy. Then, the YOLOv5s are pruned based on the convolutional kernel channel soft pruning algorithm, which greatly reduces the number of parameters in the YOLOv5-MRL model while maintaining a certain accuracy loss. Finally, in order to facilitate the dial reading, the dial external circle fitting method is proposed to calculate the dial reading using the circular angle algorithm. The experimental results on the self-built dataset show that the YOLOv5-MRL object detection model achieves a mean average precision of 96.9%, a detection speed of 5 ms/frame, and a model weight size of 5.5 MB, making it better than other advanced dial reading models.