It is reported recently that legitimate libraries are repackaged for propagating malware. An in-depth analysis of such potentially-harmful (PhaLibs), however, has never been done before, due to the challenges in identifying those whose code can be unavailable online (e.g., removed from public repositories, spreading underground, etc.). Particularly, an iOS app, library it integrates cannot trivially recovered its binary and analyzed by any publicly available anti-virus (AV) systems. In this...

LoRA (Low-Rank Adaptation) has achieved remarkable success in the parameter-efficient fine-tuning of large models. The trained matrix can be integrated with base model through addition or negation operation to improve performance on downstream tasks. However, unauthorized use LoRAs generate harmful content highlights need for effective mechanisms trace their usage. A natural solution is embed watermarks into detect misuse. existing methods struggle when multiple are combined applied, as...

The lichen genus Solorina exhibits significant morphological and chemical variations between species. Recent molecular studies have demonstrated that is polyphyletic, underscoring the need for a comprehensive taxonomic revision. Phylogenetic analyses employing Bayesian methods Maximum Likelihood approaches based on three loci (nrITS, nrLSU, mtSSU) revealed species of segregate into two distinct clades. first clade includes characterized by bright orange lower surfaces contain secondary...

As the technology for offshore wireless transmission and collaborative innovation in unmanned ships continues to mature, research has been gradually carried out various countries on methods of compressing transmitting perceptual video while driving remotely. High Efficiency Video Coding (H.265/HEVC) played an extremely important role field Unmanned Aerial Vehicle (UAV) autopilot, as one most advanced coding schemes, its performance visual sensor is excellent. According characteristics...

Maps are indispensable foundation for autonomous driving to provide prior knowledge of road structures. Constructing maps in a crowdsourcing manner has been research hotspot order achieve lower map production cost and higher update frequency. Existing pipelines mapping construct scalar on-cloud vectorization. This leads not only more demand storage space uploading bandwidth, but also the waste computing resources because vectorization is doing similar work as on-vehicle scene perception. To...

Smartphones have become common tools in people's daily life. Lots of popular applications (e.g., social network applications) immigrated from traditional computers to smartphones. With the growth apps, attacks on these apps are also increasing. One serious attack adds malicious payloads or advertisements legitimate apps. These modified called repackaged share similar functionalities with original which makes them easily spread. To mitigate this attack, we embed watermarks into Android...

Crops are greatly affected by the temperature of farmland surface during their growing period. It is feasible to investigate growth status crops based on information. For serving research crop status, component (e.g. vegetation and soil) in need be obtained. In this study, an unmanned aerial vehicle (UAV) measurement system with a thermal infrared (TIR) imager charge-coupled device (CCD) camera assembled applied measure brightness temperatures surface. The target areas were photographed UAV...

The aim of this article is to present a digital shipboard video stabilization method stabilize the unstable due unwanted camera shakes and jitters caused by motion vibration ships. proposed algorithm consists estimation, filtering compensation. idea estimation detecting corner points using adaptive Harris algorithm, then are tracked optical flow determine motion. In addition, Fourier series curve fitting designed cope with randomness After filtering, image sequence moved compensate according...

Deep learning has been widely used in fields such as natural language processing, computer vision, and driverless vehicles, leading a new wave of artificial intelligence. advances however have also to create technologies that pose potential threats national security, social stability, personal privacy. For example, deepfakes recently attracted widespread attention worldwide, which could generate seemingly realistic fake images, audio video content. This article introduces the background...

The High Efficiency Video Coding Standard (HEVC) is one of the most advanced coding schemes at present, and its excellent performance highly suitable for application in navigation field with limited bandwidth. In recent years, development emerging technologies such as screen sharing remote control has promoted process realizing virtual driving unmanned ships. order to improve transmission efficiency during sharing, HEVC proposes a new extension scheme content (HEVC-SCC), which based on...

Attackers like to install trojans in a target system control it. However, it becomes more and difficult deceive user into installing such trojans. One reason is that antivirus software uses strict policies on the first run of unknown software. The other users also become cautious. Some attackers try find vulnerabilities evade users. But not easy suitable because they are usually patched short time. In this paper, we present new type threat called vulnerability-based backdoor (VBB). It...

Recently, numerous highly-valuable Deep Neural Networks (DNNs) have been trained using deep learning algorithms. To protect the Intellectual Property (IP) of original owners over such DNN models, backdoor-based watermarks extensively studied. However, most fail upon model extraction attack, which utilizes input samples to query target and obtains corresponding outputs, thus training a substitute input-output pairs. In this paper, we propose novel watermark IP models against extraction, named...

Recent years have witnessed tremendous success in Self-Supervised Learning (SSL), which has been widely utilized to facilitate various downstream tasks Computer Vision (CV) and Natural Language Processing (NLP) domains.However, attackers may steal such SSL models commercialize them for profit, making it crucial verify the ownership of models.Most existing protection solutions (e.g., backdoorbased watermarks) are designed supervised learning cannot be used directly since they require that...

This publication has been retracted by the Editor due to identification of non-original figure images and manuscript content that raise concerns regarding credibility originality study manuscript.Reference:Ying-Jun Zhang, He Huang, Yu Liu, Bin Kong, Guangji Wang. MD-1 Deficiency Accelerates Myocardial Inflammation Apoptosis in Doxorubicin-Induced Cardiotoxicity Activating TLR4/MAPKs/Nuclear Factor kappa B (NF-kappaB) Signaling Pathway. Med Sci Monit, 2019; 25: 7898-7907. DOI: 10.12659/MSM.919861

Crop monitoring can promptly reflect the growth status of crops. However, conventional methods monitoring, although simple and direct, have limitations such as destructive sampling, reliance on human experience, slow detection speed. This study estimated fresh weight lettuce (Lactuca sativa L.) in a plant factory with artificial light based three-dimensional (3D) reconstruction technology. Data from different stages were collected training dataset, while data forms used validation dataset....

This paper introduces a cooperative computation model of services called CCM which supports data flow and channel based service computation. The meta-model is described with MOF (meta-object facility). comes from Kahn process network (KPN) model. KPN on dataflow commonly used for describing set processes that communicate each other by means dataflow. In this paper, we use to input output flows interactions among services. uses description language CCML describe including operations,...

Electric cars, as an alternative to fuel are growing rapidly in number and help reduce carbon dioxide emissions build green cities. However, due the appearance similarities between electric it is still a challenge count proportion of cars on road. Considering that thermal infrared (TIR) imaging technology can obtain temperature characteristics well work at night, this paper proposes new solution identify Unmanned aerial vehicle (UAV)-based TIR images. First, semiautomated labelling method...