This article delves into cloud computing technologies, addressing the challenges and concerns that arise in this domain, specifically focusing on resource allocation. We investigate and compare three optimization algo...
详细信息
Recent advancements in artificial intelligence (AI) and machine learning (ML) have significantly enhanced various fields, including intelligent transportation systems (ITS). A notable area of progress is vehicle class...
详细信息
Recent advancements in artificial intelligence (AI) and machine learning (ML) have significantly enhanced various fields, including intelligent transportation systems (ITS). A notable area of progress is vehicle classification, crucial for improving road safety and traffic management. Traditional vehicle classification methods often struggle with accuracy and efficiency in complex conditions. This study introduces a novel method using Neuro-Evolutionary Algorithms (NEAs) to optimize vehicle classification, combining neural networks with evolutionary computation for robust framework design and parameter optimization. NEAs adaptively refine neural network architectures, particularly enhancing their performance in diverse driving scenarios. Implemented in Google Colab, our NEA-optimized models demonstrated a remarkable classification accuracy of 98.35%, outperforming traditional and contemporary methods. This approach not only advances vehicle classification accuracy but also sets the stage for future developments in ITS, promoting safer, more efficient mobility.
This study investigates the application of deep learning models, including CNN, ResNet50, and VGG16, for the early detection of skin cancer. Utilizing a dataset of 3,307 images, the models were trained to classify ski...
详细信息
In a network design, the control plane and the data plane are separated by the architectural concept of software defined networking (SDN). A centralised controller that serves as the only point of control for the whol...
详细信息
Event cameras are increasingly popular in robotics due to beneficial features such as low latency, energy efficiency, and high dynamic range. Nevertheless, their downstream task performance is greatly influenced by th...
详细信息
ISBN:
(纸本)9798350377712;9798350377705
Event cameras are increasingly popular in robotics due to beneficial features such as low latency, energy efficiency, and high dynamic range. Nevertheless, their downstream task performance is greatly influenced by the optimization of bias parameters. These parameters, for instance, regulate the necessary change in light intensity to trigger an event, which in turn depends on factors such as the environment lighting and camera motion. This paper introduces feedback control algorithms that automatically tune the bias parameters through two interacting methods: 1) An immediate, on-the-fly fast adaptation of the refractory period, which sets the minimum interval between consecutive events, and 2) if the event rate exceeds the specified bounds even after changing the refractory period repeatedly, the controller adapts the pixel bandwidth and event thresholds, which stabilizes after a short period of noise events across all pixels (slow adaptation). Our evaluation focuses on the visual place recognition task, where incoming query images are compared to a given reference database. We conducted comprehensive evaluations of our algorithms' adaptive feedback control in real-time. To do so, we collected the QCR-Fast-and-Slow dataset that contains DAVIS346 event camera streams from 366 repeated traversals of a Scout Mini robot navigating through a 100 meter long indoor lab setting (totaling over 35km distance traveled) in varying brightness conditions with ground truth location information. Our proposed feedback controllers result in superior performance when compared to the standard bias settings and prior feedback control methods. Our findings also detail the impact of bias adjustments on task performance and feature ablation studies on the fast and slow adaptation mechanisms.
The proceedings contain 24 papers. The special focus in this conference is on New Trends in Model and Data Engineering. The topics include: Record Linkage for Auto-tuning of High Performance computingsystems;protecti...
ISBN:
(纸本)9783030876562
The proceedings contain 24 papers. The special focus in this conference is on New Trends in Model and Data Engineering. The topics include: Record Linkage for Auto-tuning of High Performance computingsystems;protecting Sensitive Data in Web of Data;COVID-DETECT: A Deep Learning Based Approach to Accelerate COVID-19 Detection;time Insertion Functions;static Checking Consistency of Temporal Requirements for control Software;visual Language for Device Management in Telecommunication Product Line;Using Process-Oriented Structured Text for IEC 61499 Function Block Specification;the DibiChain Protocol: Privacy-Preserving Discovery and Exchange of Supply Chain Information;towards a Resource-Aware Formal Modelling Language for Workflow Planning;Development of Critical systems with UML/OCL and FoCaLiZe;systematic Literature Review of Methods for Maintaining Data Integrity;medical Data Engineering – Theory and Practice;querying Medical Imaging Datasets Using Spatial Logics (Position Paper);evaluation of Anonymization Tools for Health Data;usages of the ContSys Standard: A Position Paper;systematic Assessment of Formal Methods Based Models Quality Criteria;deriving Interaction Scenarios for Timed Distributed systems by Symbolic Execution;energy Efficient Real-Time Calibration of Wireless Sensor Networks for Smart Buildings;edge-to-Fog Collaborative computing in a Swarm of Drones;Coverage Maximization in WSN Deployment Using Particle Swarm Optimization with Voronoi Diagram;EPSAAV: An Extensible Platform for Safety Analysis of Autonomous Vehicles;bridging Trust in Runtime Open Evaluation Scenarios.
UAV swarms have attracted much attention for post-disaster search and rescue, pollution monitoring and traceability, etc., where distributed scheduling is required to arrange careful tasks and time quickly. The market...
详细信息
ISBN:
(纸本)9781665491907
UAV swarms have attracted much attention for post-disaster search and rescue, pollution monitoring and traceability, etc., where distributed scheduling is required to arrange careful tasks and time quickly. The market-based methods are widely favored but they rely on the environmentally influenced communication network to complete negotiation, while the onboard computing of UAV is robust and redundant. This paper proposes a distributed scheduling method for networked UAV swarm based on computing for communication, which trades a modest increase in computing for a significant decrease in communication. First, by analyzing the task removal strategies of two representative methods, the consensus-based bundle algorithm (CBBA) and performance impact (PI) algorithm, a new removal strategy is proposed, which expands the exploration of the bundle and can potentially reduce communication rounds. Second, the proposed task-related optimization method can extract task conflict nodes from the native communication protocol, and use the sampling and estimation strategies to resolve task conflicts in advance. Third, historical bids are cleverly used to infer others' locations, which is necessary for task-related optimization. Fourth, to verify the algorithm in real communication, a hardware-in-the-loop (HIL) ad-hoc network simulation system is constructed, which uses real network protocols and simulated channel transmissions. Finally, the HIL Monte Carlo simulation results show that, compared with CBBA and PI, the proposed method can significantly reduce the number of communication rounds and the total scheduling time, without increasing the communication protocol overhead and loss of optimization.
In this article, for nonlinear continuous non-parameterized (NCNP) systems, an adaptive iterative learning control (ILC) algorithm, which can adjust the adaptive parameters in both iteration-domain and time-domain, is...
详细信息
In this article, for nonlinear continuous non-parameterized (NCNP) systems, an adaptive iterative learning control (ILC) algorithm, which can adjust the adaptive parameters in both iteration-domain and time-domain, is first proposed to track different reference trajectories repetitively over a finite time interval. As the NCNP system is required to asymptotically track reference trajectory in infinite time-domain, by virtue of partitioning the reference trajectory and system signals with a fixed time interval, the proposed adaptive ILC controller is then extended to handle the asymptotic tracking issue in infinite time-domain. Therefore, a unified adaptive control approach is practically presented for NCNP systems to track reference trajectories in different domains (the infinite iteration-domain and the infinite time-domain). A prominent feature of the unified adaptive control approach is that the unknown control gain matrices in the NCNP systems are assumed to be invertible only. As a result, the general requirement in conventional adaptive control and adaptive ILC that the control gain matrices of plants are real symmetric and positive-definite (or negative-definite) is greatly relaxed. In addition, only three adaptive variables are designed for adjusting or updating in the proposed unified adaptive control approach such that the structure of controller is very simple and the memory space for computing is saved.
In response to the increasing complexity of distributed embedded systems, self-organizing systems have emerged. One such system is organic computing, which draws inspiration from biological organisms. It improves the ...
详细信息
This paper represents on intelligent adaptive control of magnetic controlled growing rods (MGCR) which used in the treatment of early-onset scoliosis. With intelligent adaptive control theory system can be controlled ...
详细信息
暂无评论