Artificial intelligence (AI) has smoothly penetrated in a number of monitoring and control applications including agriculture. However, research efforts toward low-power sensing devices with fully functional AI on board are still fragmented. In this article, we present an embedded system enriched the AI, ensuring continuous analysis situ prediction growth dynamics plant leaves. The solution is grounded graphics processing unit (GPU) able to run neural network-based board. We use recurrent...

Artificial Intelligence (AI) has been recently applied to a number of sensing scenarios for realizing the prediction, control and/or recognition tasks. However, its integration embedded systems is still limited. We propose low-power system with AI on board special focus application in agriculture. For this reason we designed Convolutional Neural Network (CNN) which achieves 83% average Intersection over Union (IoU) score test dataset and 97% seeds accuracy validation dataset. The proposed...

The Hogweed of Sosnowskyi (lat. Heracleum sosnówskyi) is poisonous for humans, dangerous farming crops, and local ecosystems. This plant fast-growing has already spread all over Eurasia: from Germany to the Siberian part Russia, its distribution expands year-by-year. In-situ detection this harmful a tremendous challenge many countries. Meanwhile, there are no automatic systems localization hogweed. In article, we report on an approach fast accurate includes Unmanned Aerial Vehicle (UAV) with...

Aerial drones can be used for a number of monitoring and control applications. Most existing drone platforms are quite primitive in terms body-machine interface. They usually variation hand-held remote controller or ground station. However, line-of-sight scenarios it would more convenient to use the human gestures voice control. In this work, we present an approach instantaneous based on gestures. The proposed solution includes wearable sensors embedded artificial intelligence. We microphone...

Aircraft perform flight in multiple regimes with different speeds, Angles of Attack (AoA), sideslip angles, and at altitudes. Designers usually choose the airfoil having best performance for cruise mode only or being able to stay suboptimal all regimes. It leads a reduction maximum lift-to-drag ratio certain regimes, as well deterioration overall performance. That is why adaptive wing its ability optimal any promising technology which could significantly improve maneuverability aircraft...

The number of real-world Internet Things (IoT) deployments continuously and steadily increases while the capabilities single IoT devices cannot yet be exploited for purpose Artificial Intelligence. Indeed, computation complexity energy consumption are constraining requirements development implementation truly intelligent with AI. In this paper we present an experimental study where perform observation time execution, overheating, Central Processing Unit (CPU) load power using embedded system...

The article deals with a study of web-crawler behaviour on different websites. A classification web-robots, information gathering tools and their detection methods are provided. Well-known scrapers analyzed the base large web-server log set. Experimental results demonstrate that web-robot can be distinguished from human by feature analysis. research used as basis for comprehensive intrusion prevention system development.

In this article, we address the problem of hogweed detection using a drone equipped with red, green, blue (RGB) and multispectral cameras. We study two approaches: 1) offline running on orthophoto area scanned within mission 2) real-time scanning from frame stream directly edge device performing flight mission. show that by fusing information an additional camera installed drone, there is opportunity to boost quality, which can then be preserved even single RGB setup introduction convolution...

The eSports industry has greatly evolved within the last decade. Professional players have contracts with teams, perform daily routine training activities, and participate in tournaments. At same time, similar to typical sportsmen, athletes suffer from injuries that significantly reduce their performance or even prevent them participating process Also, it is not a trivial task evaluate athlete's when hiring new team member. In this study, we address these two issues. We use EEG recording...

The rapid growth of technology tends to improve human's life, save time and impose a rhythm human day activity. More than ever, the world is facing tremendous difficulties acquire, ingest, process, visualize information. This how big companies like AWS Amazon, Microsoft Azure, Google Cloud Alibaba are coming into game by providing cloud computing remediate such difficulties. To help companies, organizations, corporates users store huge amounts data, Service Providers provide storage in their...

Abstract In this paper, we present a study of web-crawlers behavior depending on the web resource. We provide simulation model web-crawler which is necessary for robots detection techniques and dataset generation methods based machine learning. analyze differences among humans, common crawlers, malicious crawlers demonstrate that their models can be used analysis. show behave similar to simulated obtain datasets traffic patterns further protection against unethical crawling. Our results...

Unmanned Aerial Vehicles (UAV) have been recently applied in several various civilian applications. Based on this, there is a growing need for intuitive UAV control interfaces. In this work, we report the Body-Machine Interface (BMI), helping human operator to quadrotor through gesture commands. We perform motion capture wearable sensors and Kalman filter reduce noise. For command recognition, designed Recurrent Neural Network recognizing gestures within 65 ms. orientation estimation,...

Leaf area and biomass are important morphological parameters for in situ plant monitoring since a leaf is vital perceiving capturing the environmental light as well represents overall development. The traditional approach measurements destructive requiring manual labor may cause damages plants. In this work, we report on AI-based assessing predicting biomass. proposed able to estimate predict plants at early stage of growth non-destructive way. For reason equip an industrial greenhouse...

These days information systems that use asymmetric cryptography or public key are widely spread all over the world. The reason of this fact is a rather high reliability existing algorithms. In terms security, we can distinguish following important tasks: ensuring availability information, its authenticity and confidentiality. All these problems be solved with help electronic digital signature (hereinafter-EDS). It ensures integrity confidentiality, origin. EDS used in various commercial...

Mixed Reality (MR) is a promising technology in the scope of Internet Things for number visual recognition applications including professional training, medical and geodesy. However, state-of-the-art approaches require extra facilities visualization positioning virtual objects real world. In this work, we propose approach video stream augmentation which relies solely on referenced to absolute reference frame. This can be applied dynamic platform with six dimensions freedom. For testing...

Aircrafts perform flight in multiple regimes with different speeds, Angles of Attack (AoA), sideslip angles and altitudes.Designers usually choose the airfoil best performance for cruise mode only or which stays suboptimal all regimes.It leads to reduction maximum lift-to-drag ratio certain regime as well overall performance.That is why adaptive wing, can stay optimal any promising technology, could significantly improve maneuverability aircraft during flight.This work demonstrates wing...