Transfer learning (TL) has been successfully applied to many real-world problems that traditional machine (ML) cannot handle, such as image processing, speech recognition, and natural language processing (NLP). Commonly, TL tends address three main of learning: (1) insufficient labeled data, (2) incompatible computation power, (3) distribution mismatch. In general, can be organized into four categories: transductive learning, inductive unsupervised negative learning. Furthermore, each...

Unmanned aircraft systems (UAS), or unmanned aerial vehicles (UAVs), often referred to as drones, have been experiencing healthy growth in the United States and around world. The positive uses of UAS potential save lives, increase safety efficiency, enable more effective science engineering research. However, are subject threats stemming from increasing reliance on computer communication technologies, which place public safety, national security, individual privacy at risk. To promote safe,...

The Internet of Things (IoT) is becoming an indispensable part everyday life, enabling a variety emerging services and applications. However, the presence rogue IoT devices has exposed to untold risks with severe consequences. first step in securing detecting identifying legitimate ones. Conventional approaches use cryptographic mechanisms authenticate verify devices' identities. protocols are not available many systems. Meanwhile, these methods less effective when can be exploited or...

The advancement of the micro-electro-mechanical sensory industry and open source autopilot stacks, have dramatically reduced cost difficulty making highly maneuverable UAVs. Such ease in flying drones has caused concerns about privacy, public safety, security. One major threats is inadequate control small over sensitive areas. In this research, we address problem through a dual approach: detection eviction. We propose distributed system to identify appearance approximate position unwelcome...

Medical image processing is one of the most important topics in Internet Things (IoMT). Recently, deep learning methods have carried out state-of-the-art performances on medical imaging tasks. In this paper, we propose a novel transfer framework for classification. Moreover, apply our method COVID-19 diagnosis with lung Computed Tomography (CT) images. However, well-labeled training data sets cannot be easily accessed due to disease's novelty and privacy policies. The proposed has two...

The Internet of Things (IoT) provides applications and services that would otherwise not be possible. However, the open nature IoT make it vulnerable to cybersecurity threats. Especially, identity spoofing attacks, where an adversary passively listens existing radio communications then mimic legitimate devices conduct malicious activities. Existing solutions employ cryptographic signatures verify trustworthiness received information. In prevalent IoT, secret keys for cryptography can...

Bitcoin combines a peer-to-peer network and cryptographic algorithm to implement distributed digital currency system, which keeps all transaction history on public blockchain. Since transactions recorded the blockchain are everyone, users face threat of leaking financial privacy. Many analysis deanonymization approaches have been proposed link records real identities. To eliminate this threat, we present an unlinkable coin mixing scheme that allows mix their bitcoins without trusting third...

Wireless communication involving unmanned aerial vehicles (UAVs) is expected to play an important role in future wireless networks. However, different from conventional terrestrial systems, UAVs typically have rather limited onboard energy on one hand, and require additional flying consumption the other hand. This renders energy-efficient UAV with smart expenditure of paramount importance. In this paper, via extensive flight experiments, we aim firstly validate recently derived theoretical...

Deep learning (DL) has been utilized pervasively in the Internet of Things (IoT). One typical application DL IoT is device identification from wireless signals, namely, noncryptographic (NDI). However, components NDI systems have to evolve adapt operational variations, such a paradigm termed as incremental (IL). Various IL algorithms proposed and many them require dedicated space store increasing amount historical data, therefore, they are not suitable for or mobile applications. Besides,...

Autonomous ultrasound (US) acquisition is an important yet challenging task, as it involves interpretation of the highly complex and variable images their spatial relationships. In this work, we propose a deep reinforcement learning framework to autonomously control 6-D pose virtual US probe based on real-time image feedback navigate towards standard scan planes under restrictions in real-world scans. Furthermore, confidence-based approach encode optimization quality process. We validate our...

The sewage treatment system based on particle swarm optimization algorithm to improve energy efficiency is analyzed. Firstly, this paper takes and activated sludge method as the research object, with help of experimental analysis method, has achieved good results. From results system, it plays an important role in realizing effective improving economic social benefits treatment. Then, can play system. Therefore, actively promoting application Finally, optimal control solution improved...

Privacy preserving becomes an important issue in the development progress of data mining techniques. has become increasingly popular because it allows sharing privacy-sensitive for analysis purposes. So people have unwilling to share their data, frequently resulting individuals either refusing or providing incorrect data. In turn, such problems collection can affect success mining, which relies on sufficient amounts accurate order produce meaningful results. recent years, wide availability...

Benefited from cloud storage services, users can save their cost of buying expensive and application servers, as well deploying maintaining applications. Meanwhile they lost the physical control data. So effective methods are needed to verify correctness data stored at which research issues Provable Data Possession (PDP) faced. The most important features in PDP are: 1) supporting for public, unlimited numbers times verification; 2) dynamic update; 3) efficiency space computing. In mobile...

To protect sensitive information in mined data, researchers need a way to organize variety of ongoing work. The Rampart framework categorizes protection approaches and encourages interdisciplinary solutions the growing privacy problems associated with knowledge discovery from data.

The inland aquaculture environment is an artificial ecosystem, where the water quality a key factor which closely related to economic benefits of and aquatic products. Compared with marine aquaculture, normally smaller susceptible pollution, poor self-purification capacity. Considering its low cost large-scale monitoring ability, many researches have developed spectrum sensor on-board satellite platforms allow remote surface. However, there remain problems, such as image resolution, flexible...

The ubiquitous deployment of 5G New Radio (5G NR) accelerates the evolution in many fields. With enhancement NR, unmanned aerial vehicle (UAV) swarm networking can gain more flexibility, reliability, and elasticity to assist residents or workers finish missions high complexity risks remotely. beamforming NR improve accuracy flexibility connections between mobile devices. a throughput guaranteed UAV networking, remote deliver specific instructions with requirement. Further, reliable volume...

The semantic segmentation of high-resolution remote sensing images (HRRSIs) is a basic task for image processing and has wide range applications. However, the abundant texture information imaging HRRSIs lead to complex distribution ground objects unclear boundaries, which bring huge challenges HRRSIs. To solve this problem, in paper we propose an improved squeeze excitation residual network (SERNet), integrates several modules (SERMs) refine attention module (RAM). SERM can recalibrate...

Freezing the pre-trained backbone has become a standard paradigm to avoid overfitting in few-shot segmentation. In this paper, we rethink and explore new regime: {\em fine-tuning small part of parameters backbone}. We present solution overcome problem, leading better model generalization on learning novel classes. Our method decomposes into three successive matrices via Singular Value Decomposition (SVD), then only fine-tunes singular values} keeps others frozen. The above design allows...

Generative models based on Variational Autoencoders (VAEs) represent an important area of research in Controllable Text Generation (CTG). However, existing approaches often do not fully exploit the potential latent variables, leading to limitations both diversity and thematic consistency generated text. To overcome these challenges, this paper introduces a new framework Hierarchical Latent Modulation (HLM). The incorporates hierarchical space modulation module for generation embedding...