In Drone-vs-Bird Detection Challenge in conjunction with the 4th International Workshop on Small-Drone Surveillance, and Counteraction Techniques at IEEE AVSS 2021, we proposed a YOLOV5-based object detection model for small UAV classification. YOLOV5 leverages PANet neck mosaic augmentation which help improving of objects. We have combined challenge dataset one publicly available air to having complex background lighting conditions training model. The approach achieved 0.96 Recall, $0.98...

This paper presents the 4-th edition of "drone-vs-bird" detection challenge, launched in conjunction with 17-th IEEE International Conference on Advanced Video and Signal-based Surveillance (AVSS). The objective challenge is to tackle problem detecting presence one or more drones video scenes where birds may suddenly appear, taking into account some important effects such as background foreground motion. proposed solutions should identify localize scene only when they are actually present,...

Respiration rate (RR) and respiration patterns (RP) are considered early indicators of physiological conditions cardiorespiratory diseases. In this study, we addressed the problem contactless estimation RR classification RP one person or two persons in a confined space under realistic conditions. We used three impulse radio ultrawideband (IR-UWB) radars 3D depth camera (Kinect) to avoid any blind spot room ensure that at least covers monitored subjects. This article proposes subject...

The advancement of UAS (Unmanned Aircraft Systems) technologies, and the rise in potential misuse these vehicle platforms, particular small (sUAS), has highlighted demand for a robust reliable solution detection classification aircraft (commonly referred to by Drone)vs. other flying objects. Most existing research addresses this problem either extracting micro-Doppler from radar data or features visual data. But solutions do not perform well all weather conditions beyond distance. To solve...

This paper studies a generalized class of restless multi-armed bandits with hidden states and allow cumulative feedback, as opposed to the conventional instantaneous feedback. We call them lazy (LRBs) events decision making are sparser than state transition. Hence, feedback after each event is effect following transition events. The arms from maker rewards for actions dependent. needs choose one arm in interval, such that long-term reward maximized. As hidden, maintains updates its belief...

Restless multi-armed bandits are a class of discrete-time stochastic control problems which involve sequential decision making with finite set actions (set arms). This paper studies constrained restless (CRMAB). The constraints in the form time varying available variation can be either or semi-deterministic. Given arms, fixed number them chosen to played each interval. play arm yields state dependent reward. current states arms partially observable through binary feedback signals from that...

An intent modelling and inference framework is presented to assist the defense planning for protecting a geo-fence against unauthorized flights. First, novel mathematical definition of an uncrewed aircraft system (UAS) presented. The concepts critical waypoints waypoint patterns are introduced associated with motion process fully characterize intent. This consists representations UAS mission planner, used plan aircraft's sequence, as well defined protect geo-fence. It applicable autonomous,...

Object distance estimation using the monocular camera is a challenging problem in computer vision with many practical applications. Various algorithms are developed for camera; some involve traditional techniques, while others based on Deep Learning (DL). Both methods have limitations, such as requiring calibration parameters, limited range, or object of interest should be relatively large to get accurate estimation. Due these drawbacks, cannot easily generalized In this paper, we propose...

Restless multi-armed bandits(RMAB) with partially observable states have been extensively studied for scheduling in opportunistic communication systems. These RMAB models assume that when the decision maker plays a particular arm, it gathers information about system state through feedback signals. allow only one transition single interval. In this paper, we propose cumulative model, where multiple transitions occur We formulate systems and relay selection problem as feedback. of an arm is...

Respiratory rate (RR) is one of the vital signs which commonly measured by contact-based methods, such as using a breathing belt. Recently, significant research has been conducted related to contactless RR monitoring - however, majority experiments are performed in situations when subject oriented towards radar. In this research, we interested subjects who can be anywhere room. A system three impulse radio ultrawideband (IR-UWB) radars used cover whole Kinect camera that track subjects'...

The 3 V's defining big data raises the need for non-conventional computing and communication architectures its analytics processing. In this paper we evolve a new architecture based on amalgamating benefits of Crowdsourcing Cloud Computing. "Crowd-Cloud" so formed presents newer domains efficient analysis processing data. By integrating sensing capabilities mobile dynamic cloud, provides an increased functionality in handling Big Data. proposed furthermore has two-fold benefits, minimizing...

This paper introduces the multi-view Air-to-Air Simulated Drone Dataset (A2A-SDD), a comprehensive simulated drone dataset captured using AirSim <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">©</sup> . The encompasses diverse scenarios where one or two drones are pursued by to three monitoring drones. It includes five types of drones, such as DJI models and generic quadrotor model, recorded in various weather conditions environments. Both...

The recombinant adeno-associated virus (rAAV) is a viral vector technology for gene therapy that considered the safest and most effective way to repair single-gene abnormalities in non-dividing cells. However, improving titer productivity rAAV production remains challenging. first step this end effectively monitor process state variables (cell density, GLC, GLN, LAC, AMM, titer) improve control performance an enhanced productivity. current approaches monitoring are expensive, laborious,...

The problem of rested and restless multi-armed bandits with constrained availability (RMAB-CA) arms is considered. states evolve in Markovian manner the exact are hidden from decision maker. First, some structural results on value functions claimed. Following these results, optimal policy turns out to be a threshold policy. Furthermore, indexability established for both RMAB-CAs. An index formula derived model, while an algorithm provided case.

All sorts of information today is getting digitized and internet among others the easiest a very effective source host for sharing data.So, it becomes important to track, protect monitor data maintain its security, confidentiality, integrity, authentication, robustness ownership.And, techniques like steganography, watermarking cryptography help us do so.In hidden inside (like image, audio, video, text) confidentiality maintained as only sender receiver know how retrace information.In this...

Abstract Training Deep Learning (DL) models with missing labels is a challenge in diverse engineering applications. Missing value imputation methods have been proposed to try address this problem, but their performance affected Massive Proportion of Labels (MPML). This paper presents approach for handling MPML Multivariate Long-Term Time Series Forecasting. It an two-step process where interpolation (using Gaussian Processes Regression (GPR) and domain knowledge from experts) prediction...

This paper introduces a novel approach to es-timating the distance between drone and camera using deep learning techniques. The proposed method employs low-complexity convolutional neural network (CNN), called DroneRanger, analyze captured 2D image estimate observer target drones. Three types of input data for CNN regression model are investi-gated, including extended bounding box, resized box with additional size information. effectiveness is demonstrated through experi-ments conducted on...

Dependable visual drone detection is crucial for the secure integration of drones into airspace. However, accuracy significantly affected by domain shifts due to environmental changes, varied points view, and background shifts. To address these challenges, we present DrIFT dataset, specifically developed under includes fourteen distinct domains, each characterized in point synthetic-to-real data, season, adverse weather. uniquely emphasizes shift providing segmentation maps enable...

In recent years, drones have become popular for various applications, including surveillance and delivery. Detecting classifying is crucial security regulatory purposes, as well improving the efficiency of drone operations. this work, we propose a human-AI collaboration framework detection, tracking, classification using multiple sensors, such ground radar pan-tilt-zoom (PTZ) camera. Our approach combines traditional signal processing machine learning to classify tracks, deep learning-based...

Unmanned Aerial Vehicles (UAVs) have experienced remarkable progress and widespread utilization, highlighting the need for robust detection classification systems to ensure safety security. This paper presents a comprehensive study on UAV payload classification, emphasizing fusion of multiple data sources enhance accuracy reliability. A system integrating radar Pan-Tilt-Zoom (PTZ) camera is developed evaluated. Experimental results demonstrate effectiveness proposed approach, achieving > 94%...

The emergence of innovative technologies for mobile devices has enabled us to crowdsource large scale tasks these devices. This gives rise a new paradigm where clouds can be accessed by an end user data storage and processing over wireless connection. A crowd-cloud system consists central Consumer Crowd Interface several crowd-clouds. Each comprises which perform computations allotted them. paper presents mathematical formulation the total energy consumption in such including offloading,...

We consider the problem of sensor scheduling in energy constrained network. It is modeled using restless multi-armed bandits with dynamic availability arms. An arm represents and due to its dynamic. The data transmission rate depends on channel quality. Sensor a sequential decision which needs account both for evolution quality fluctuation levels nodes. When available scheduled, it yields based quality, this referred as immediate reward. two state Markov model. higher corresponds hence...