Hand gesture serves as a crucial role during the expression of sign language. Current deep learning based methods for sign language understanding (SLU) are prone to over-fitting due to insufficient sign data resource ...
详细信息
High Utility Itemset Mining (HUIM) and Frequent Itemset Mining (FIM) are two important branches in the data mining area, where Frequent Itemset Mining considers itemsets that occur in large numbers in the transaction ...
详细信息
Great progress has been made toward accurate face detection in recent ***,the heavy model and expensive computation costs make it difficult to deploy many detectors on mobile and embedded devices where model size and ...
详细信息
Great progress has been made toward accurate face detection in recent ***,the heavy model and expensive computation costs make it difficult to deploy many detectors on mobile and embedded devices where model size and latency are highly *** this paper,we present a millisecond-level anchor-free face detector,YuNet,which is specifically designed for edge *** are several key contributions in improving the efficiency-accuracy ***,we analyse the influential state-of-theart face detectors in recent years and summarize the rules to reduce the size of ***,a lightweight face detector,YuNet,is *** detector contains a tiny and efficient feature extraction backbone and a simplified pyramid feature fusion *** the best of our knowledge,YuNet has the best trade-off between accuracy and *** has only 75856 parameters and is less than 1/5 of other small-size *** addition,a training strategy is presented for the tiny face detector,and it can effectively train models with the same distribution of the training *** proposed YuNet achieves 81.1%mAP(single-scale)on the WIDER FACE validation hard track with a high inference efficiency(Intel i7-12700K:1.6ms per frame at 320×320).Because of its unique advantages,the repository for YuNet and its predecessors has been popular at GitHub and gained more than 11K stars at https://***/ShiqiYu/***:Face detection,object detection,computer version,lightweight,inference efficiency,anchor-free mechanism.
Pretrained language models leverage selfsupervised learning to use large amounts of unlabeled text for learning contextual representations of sequences. However, in the domain of medical conversations, the availabilit...
详细信息
Previous articles on unsupervised skeleton-based action recognition primarily focused on strategies for utilizing features to drive model optimization through methods like contrastive learning and reconstruction. Howe...
详细信息
VANETs are highly attractive and is used in maximum of the applications of cross-regional communication. To increase the coverage of the vehicular network, Unmanned Arial Vehicles (UAVs) are introduced, and they get c...
详细信息
VANETs are highly attractive and is used in maximum of the applications of cross-regional communication. To increase the coverage of the vehicular network, Unmanned Arial Vehicles (UAVs) are introduced, and they get connected with the satellite networks to perform heterogeneous communication. With the help of this connectivity, the communication quality of ground level to air medium is increased. Currently the vehicle usage is highly increased and as a results of communication link failure, improper resource allocation are arises whither abruptly assumes a stability about a network with that increases an energy consumption and communication delay in the heterogeneous networks. In these conditions, thus study is idea of Resource Allocation and Edge Computing for Dual Hop Communication (RAEDH) in introduced in satellite assisted UAVs enabled VANETs. The major sections of the approach are UAV assisted mobile computing, resource allocation among the vehicles and the UAVs, and dual communication among the vehicles and the *** these methods the input resources are properly allocated and that reduces the power utility and communication delay. Initially, the vehicular network is established, incorporating trusted components like TA, RSU, and CRS. Subsequently, mobile edge computing reduces energy consumption through computation offloading and optimized UAV trajectory selection. Resource allocation, facilitated by whale optimization, ensures effective utilization across vehicles. The implementation of this method is done in NS3, and the scenario is analyzed using two parameters like number of vehicles and its speed. The output parameters that remain thought-out over a performance examination stay throughput, end-to-end delay, energy efficiency, packet loss, packet delivery ratio, and routing overhead, and as well those results are compared with the earlier methods. Finally, dual-hop transmission between vehicles and UAVs enhances delivery ratio and throughput. From
This study investigates the application of deep learning,ensemble learning,metaheuristic optimization,and image processing techniques for detecting lung and colon cancers,aiming to enhance treatment efficacy and impro...
详细信息
This study investigates the application of deep learning,ensemble learning,metaheuristic optimization,and image processing techniques for detecting lung and colon cancers,aiming to enhance treatment efficacy and improve survival *** introduce a metaheuristic-driven two-stage ensemble deep learning model for efficient lung/colon cancer *** diagnosis of lung and colon cancers is attempted using several unique indicators by different versions of deep Convolutional Neural Networks(CNNs)in feature extraction and model constructions,and utilizing the power of various Machine Learning(ML)algorithms for final ***,we consider different scenarios consisting of two-class colon cancer,three-class lung cancer,and fiveclass combined lung/colon cancer to conduct feature extraction using four *** extracted features are then integrated to create a comprehensive feature *** the next step,the optimization of the feature selection is conducted using a metaheuristic algorithm based on the Electric Eel Foraging Optimization(EEFO).This optimized feature subset is subsequently employed in various ML algorithms to determine the most effective ones through a rigorous evaluation *** top-performing algorithms are refined using the High-Performance Filter(HPF)and integrated into an ensemble learning framework employing weighted *** findings indicate that the proposed ensemble learning model significantly surpasses existing methods in classification accuracy across all datasets,achieving accuracies of 99.85%for the two-class,98.70%for the three-class,and 98.96%for the five-class datasets.
This paper introduces a novel software technique to optimize thread allocation for merged and fused kernels in multi-Tenant inference systems on embedded Graphics Processing Units (GPUs). Embedded systems equipped wit...
详细信息
Combining optical and electronic systems could enable information processing that is a million times faster than existing gigahertz technology. Imagine leveraging nature’s fastest processes to power the electronics i...
详细信息
Combining optical and electronic systems could enable information processing that is a million times faster than existing gigahertz technology. Imagine leveraging nature’s fastest processes to power the electronics in semiconductor chips, quantum sensors and quantum computers. Such transformative speed would not only greatly improve the performance of technology, but unveil new vistas for fundamental science as well.
With the rise of cloud computing, multi-user scenarios have become a common setting for data sharing nowadays. The conservative security notion might not be sufficient for such a data sharing model. As a response to t...
详细信息
暂无评论