Traditional fault diagnosis methods are often inefficient and costly, hindered by limited computing power that compromises accuracy. Their lack of flexibility also poses challenges in establishing effective multi-devi...
详细信息
Traffic accidents pose a major risk to human life and have a considerable economic impact worldwide. In this study, we present the RAPID framework, an innovative approach that integrates Artificial Intelligence (AI) a...
详细信息
Designing machine learning applications can be challenging, especially software architectures for handling real-time sensor data processed by compute- and software-intensive machine learning applications. This paper a...
详细信息
ISBN:
(纸本)9798350366266;9798350366259
Designing machine learning applications can be challenging, especially software architectures for handling real-time sensor data processed by compute- and software-intensive machine learning applications. This paper answers the two research questions: "Which parts of a machine learning pipeline do novel software architectures and framework optimize?" and the associated question of "Which parts do the architect then need to focus on?". The presented experiences and experimental results suggest that novel software architectures and frameworks optimize the learning and classification part of the pipelines. Therefore, the architect in particular needs to focus on data distribution and preprocessing as these parts were observed to have an overlooked computational cost and complexity. These results are important for software architects to become better at architecting machine learning-based systems.
This paper proposes an architecture of an ARM-based remote video electronic information monitoring system. The system converts the embedded Linux terminal system memory allocation problem into a 0-1 knapsack problem. ...
详细信息
Implementing aerial transportation systems in real-world scenarios is becoming feasible. Due to the ability to dynamically adjust the distance between quadrotor and cargo, the aerial transportation system with variabl...
详细信息
High Performance computing (HPC) systems generate a large amount of unstructured/alphanumeric log messages that capture the health state of their components. Due to their design complexity, HPC systems often undergo f...
详细信息
ISBN:
(纸本)9798350347937
High Performance computing (HPC) systems generate a large amount of unstructured/alphanumeric log messages that capture the health state of their components. Due to their design complexity, HPC systems often undergo failures that halt applications (e.g., weather prediction, aerodynamics simulation) execution. However, existing failure prediction methods, which typically seek to extract some information theoretic features, fail to scale both in terms of accuracy and prediction speed, limiting their adoption in real-time production systems. In this paper, differently from existing work and inspired by current transformer-based neural networks which have revolutionized the sequential learning in the natural language processing (NLP) tasks, we propose a novel scalable log-based, self-supervised model (i.e., no need for manual labels), called time Machine1, that predicts (i) forthcoming log events (ii) the upcoming failure and its location and (iii) the expected lead time to failure. time Machine is designed by combining two stacks of transformer-decoders, each employing the self-attention mechanism. The first stack addresses the failure location by predicting the sequence of log events and then identifying if a failure event is part of that sequence. The lead time to predicted failure is addressed by the second stack. We evaluate time Machine on four realworld HPC log datasets and compare it against three state-of-the-art failure prediction approaches. Results show that time Machine significantly outperforms the related works on Bleu, Rouge, MCC, and F1-score in predicting forthcoming events, failure location, failure lead-time, with higher prediction speed.
Traffic sign recognition is crucial for the safe and efficient operation of autonomous vehicles. While previous research has primarily focused on traffic sign recognition in foreign countries, these studies often face...
详细信息
ISBN:
(纸本)9798350352368
Traffic sign recognition is crucial for the safe and efficient operation of autonomous vehicles. While previous research has primarily focused on traffic sign recognition in foreign countries, these studies often face limitations such as differing traffic sign designs, language barriers in textual information, and varying environmental conditions. In this paper, we propose a traffic sign detection and recognition system tailored for Malaysia, utilizing Convolutional Neural Networks (CNNs) and Optical Character Recognition (OCR). In this paper, we propose a traffic sign detection and recognition system utilizing You Only Look Once (YOLO) V8 for object detection and EasyOCR to process textual information on selected traffic signs. Our system achieves a mean Average Precision (mAP) of 0.824 and an average processing time of 1.2 seconds per frame, which is comparable to existing literature. Furthermore, the complexity of our method is significantly reduced, enhancing its potential for real-time processing applications, as evidenced by its efficient processing time.
The growth of the Internet of Vehicles (IoV) has introduced new challenges and opportunities. Among the most crucial considerations is ensuring the safety of passengers and the environment. Connected vehicles offer nu...
详细信息
ISBN:
(纸本)9798350333398
The growth of the Internet of Vehicles (IoV) has introduced new challenges and opportunities. Among the most crucial considerations is ensuring the safety of passengers and the environment. Connected vehicles offer numerous benefits, but they are also at risk of fire incidents caused by various factors such as electrical failures, fuel leaks, and collisions. These events can result in devastating outcomes, including property loss, injury, and even loss of life. The conventional fire detection systems employed in vehicles are large, expensive, and consume a considerable amount of power, making them incompatible with the resource-limited environment of the IoV. The present work overcomes these limitations with the introduction of FlameNet, a custom-designed neural network for fire detection. FlameNet not only outperforms existing solutions but also boasts a lightweight design, which contributes to its high computational efficiency, allowing it to run smoothly on low-cost embedded devices with a frame rate of 28 frames per second. Accuracy, recall, precision, and F-measure were used to assess the model's efficiency on both industry-standard fire datasets and a custom-built test set. The results showed that FlameNet performed well on both datasets, with its performance being better on the standard fire test dataset due to its limited image diversity. The performance of the model is encouraging, and the IoT functionality allows immediate visual feedback and a fire alarm in the event of an emergency.
The embedded Graphics Processing Unit (GPU) module, which includes both Central Processing Unit (CPU) and GPU processors, can be easily integrated into radar systems, offering high performance and flexibility. Phased ...
详细信息
We propose fitting a distributed realtime Ethernet (RTE) device with more than one RTE communication stack. We target this proposal for increasing flexibility for the manufacturing and system commissioning phases as ...
详细信息
暂无评论