this paper presents a novel method to find optimal Bidirectional Long-Short Term Memory Neural Network (Bi-LSTM) using Bayesian Optimisation method for vehicle trajectory classification. We extend our previous approac...
详细信息
ISBN:
(纸本)9783031457241;9783031457258
this paper presents a novel method to find optimal Bidirectional Long-Short Term Memory Neural Network (Bi-LSTM) using Bayesian Optimisation method for vehicle trajectory classification. We extend our previous approach to be able to classify a larger number of vehicle trajectories collected from different sources in a single Bi-LSTM network. We also explored the use of deep learning visual explainability by highlighting the parts of the activity (or trajectory) contribute to the classification decision of the network. In particular, Qualitative Trajectory Calculus (QTC), spatio-temporal calculus, method is used to encode the relative movement between vehicles as a trajectory of QTC states. We then develop a Bi-LSTM network (called VNet) to classify QTC trajectories that represent vehicle pairwise activities. Existing Bi-LSTM networks for vehicle activity analysis are manually designed without considering the optimisation of the whole architecture nor its trainable hyperparameters. therefore, we adapt Bayesian Optimisation method to search for an optimal Bi-LSTM architecture for classifying QTC trajectories of vehicle interaction. To test the validity of the proposed VNet, four datasets of 8237 trajectories of 9 unique vehicle activities in different traffic scenarios are used. We further compare our VNet model's performance withthe state-of-the-art methods. the results on the combined dataset (accuracy of 98.21%) showed that the proposed method generates light and most robust Bi-LSTM model. We also demonstrate that Activation Map is a promising approach for visualising the Bi-LSTM model decisions for vehicle activity recognition.
this paper studies a mobile edge computing system where two solar-powered nodes (i.e., a relay and an intelligent reflecting surface (IRS)) assist a user node in task offloading to an access point. To save the long-te...
详细信息
ISBN:
(纸本)9783031705069;9783031705076
this paper studies a mobile edge computing system where two solar-powered nodes (i.e., a relay and an intelligent reflecting surface (IRS)) assist a user node in task offloading to an access point. To save the long-term energy consumption at the user, a novel protocol is first proposed so that the system can adaptively select the operating modes. then, based on this protocol, the optimization problem of the system is formulated to minimize the energy consumption of task offloading and computing at the user by optimizing the system operation modes and the resource allocation in each mode, subject to the battery energy states of the IRS and the relay withthe energy causality constraints. the problem is solved using the Lyapunov optimization framework and an alternating optimization algorithm. Simulation results show that the proposed system optimization scheme can save 70%-95% of energy consumption as compared to the baseline schemes.
this paper proposes a design of a 1.5-V 500-MHz fully-differential CMOS sample-and-hold (S/H) circuit using double-sampling for ADCs. In contrast to a traditional closed-loop S/H circuit, a double-sampled S/H circuit ...
详细信息
Load forecasting is an essential part in power system management and planning. In recent years, more and more researchers started to focus on the behavior of terminal customers, trying to improve the accuracy of load ...
详细信息
Drug response prediction plays a crucial role in precision medicine, such as cancer analysis and treatment. Due to the uncertainty of drug efficacy and the heterogeneity of cancer, predicting drug response in vitro is...
详细信息
the heart is one of the most essential organs in the human body, working as the major pump for the blood that passes through the veins of the body, transporting energy and nutrients needed for the body to function. Ho...
详细信息
Solar minimum has occurred multiple times since recorded by humans. In this study, we utilize four well-known solar minimums as known data to predict the upcoming solar minimum. We employ the 3-input Weng-Zhang (WZ) a...
详细信息
Cloud Network Slicing (CNS) is a concept that describes a mechanism to provide computing, networking, and storage as a virtual slice entity, enabling new approaches to applications and structuring resources at the edg...
详细信息
ISBN:
(纸本)9798350399806
Cloud Network Slicing (CNS) is a concept that describes a mechanism to provide computing, networking, and storage as a virtual slice entity, enabling new approaches to applications and structuring resources at the edge of the network. In this paper, the architecture defined in the NECOS Project is adopted, and the functions for creating CNS in resource-constrained edge devices were designed and implemented. the implementation was evaluated on Single Board computers (SBCs), using lightweight virtualization solutions (microservices) and the results achieved show that it is possible to instantiate CNSs on that hardware, however, also show some limitations of multiple slice support on resource-constrained devices.
A number of attempts have already been implemented formally to solve road traffic congestion. However, the objective strategy type of road traffic engineering could not be proven truly. Try and error is one inefficien...
详细信息
BacaDisleksia is an application specifically designed for children with dyslexia learning to read. the application aims to facilitate dyslexic children and ease their reading by carefully considering the Human-Compute...
详细信息
暂无评论