Withthe evolution of robotic systems, unmanned aerial vehicles (UAV) have become a target of interest for domains such as computer vision (CV) and artificial intelligence (AI), contributing to a variety of applicatio...
详细信息
ISBN:
(纸本)9781450398541
Withthe evolution of robotic systems, unmanned aerial vehicles (UAV) have become a target of interest for domains such as computer vision (CV) and artificial intelligence (AI), contributing to a variety of applications for surveillance, transportation and many more. A very hot topic that is the playground of the proposed benchmark is visual human tracking in images acquired by a camera mounted on a UAV. this target application troubles CV and deep learning (DL) research community in recent years and it has created serious demands for visual tracking algorithms. Some of the most important demands are high performance under hard visual tracking conditions and deployment in edge devices with limited computation resources. these two challenges are the main motivation of the presented paper, where 37 tracking algorithms have been benchmarked in visual object tracking (VOT) images. For each tracking algorithm two metric categories, relative to detection performance and hardware resources consumption, have been considered. the objective of the proposed paper is to highlight the most lightweight and high performance tracking algorithms for usage in UAV based applications.
To solve the problem that it is difficult to judge the multiple aircraft behaviors through the corresponding trajectory information, all kinds of which are similar in the airport scene video monitoring system, multi-a...
详细信息
ISBN:
(数字)9798350389418
ISBN:
(纸本)9798350389425
To solve the problem that it is difficult to judge the multiple aircraft behaviors through the corresponding trajectory information, all kinds of which are similar in the airport scene video monitoring system, multi-aircraft behavior recognition method for airport scene based on ASTERIX data is proposed. Firstly, ASTERIX data is used to locate the aircraft in the video by the perspective transformation and the historical trajectory information is used to predict the location of aircraft at the next moment, through which can extract the movement area of each aircraft. Namely, the multi-aircraft behavior recognition task is transformed into multiple single-aircraft behavior recognition tasks. Finally, the behavior recognition is completed by the fusion convolution with multi-time scales method in the movement region of each aircraft. the experiment based on a self-made dataset containing various aircraft behaviors at typical airport shows that the proposed method meets the behavior recognition task of multiple targets in the airport scene video monitoring system.
Employing Enterprise Survey data that comprises 1276 firms from four Central Asian countries, this study focuses on the impact of digitalization as well as human capital on firms’ productivity. We have found that dig...
ISBN:
(纸本)9781450399050
Employing Enterprise Survey data that comprises 1276 firms from four Central Asian countries, this study focuses on the impact of digitalization as well as human capital on firms’ productivity. We have found that digitalization improves firms’ productivity by 44-52 percent on average, statistically significant at 1%. Moreover, analysis shows that the impact is heterogenous, with large and statistically significant effects for firms in retail, nonmetal, and, especially, textile industries. these results imply that digitalization is crucial in improving firms’ productivity and the government could implement policies that encourage and help firms withthis regard.
the paper presents an innovative car social network service that allows getting in touch with a driver through his individual identification based on the car license plate. the Car Social Network architecture is based...
详细信息
ISBN:
(纸本)9781665426053
the paper presents an innovative car social network service that allows getting in touch with a driver through his individual identification based on the car license plate. the Car Social Network architecture is based on a unified cloud communication platform that enables the activation of multimedia communication sessions withthe car *** application contexts range from road safety to the reduction of noise pollution in cities.
the proceedings contain 6 papers. the topics discussed include: inspiring healthy habits: data science at WW;utilizing collaborative filtering to recommend opportunities for positive affect in daily life;personalized,...
the proceedings contain 6 papers. the topics discussed include: inspiring healthy habits: data science at WW;utilizing collaborative filtering to recommend opportunities for positive affect in daily life;personalized, health-aware recipe recommendation: an ensemble topic modeling based approach;rethinking hearing aids as recommender systems;evolutionary approach for ‘healthy bundle’ wellbeing recommendations;and an evaluation of recommendation algorithms for online recipe portals.
Industry 4.0 currently prepares a major shift towards extreme flexibility into production lines management. Digital Twins are one of the key enabling technologies for Industry 4.0. However, the interoperability gap am...
详细信息
Industry 4.0 currently prepares a major shift towards extreme flexibility into production lines management. Digital Twins are one of the key enabling technologies for Industry 4.0. However, the interoperability gap among digital representation of Industry 4.0 assets is still one of the obstacles to the development and adoption of digital twins. If the Asset Administration Shell (AAS), the standard proposed to represent the I4.0 components, caters for syntactic interoperability, a more semantic kind of interoperability is deeply needed to develop flexible and adaptable production lines. In our work, we overcome the limitation of current syntactic-only resource matching algorithms by implementing semantic interoperability based on ontologies i.e., by transforming AAS-based plant models into MaRCO (Manufacturing Resource Capability Ontology) instances and then query the expanded ontology to find the needed resources. this article presents this ontology-based approach as the first step towards the design and implementation of an automated I4.0 flexible plant supervision and control system based on model-driven engineering (MDE) within the “Papyrus for Manufacturing” toolset. We show how an MDE approach can aggregate around digital twin modeling tools from the Papyrus platform both I4.0 technologies and AI (Knowledge Representation and Reasoning) tools. Our platform aligns modeling and ontological elements to get the best of both worlds. this method has two main advantages: (1) to provide semantic descriptions for digital twin models, (2) to complement model-driven engineering tools with automated reasoning. this paper showcases this approach through a robotic cell use case.
the emergence of 6G networks heralds a transformative era in network slicing, facilitating tailored service delivery and optimal resource utilization. Despite its promise, network slice optimization heavily relies on ...
详细信息
ISBN:
(数字)9798350369588
ISBN:
(纸本)9798350369595
the emergence of 6G networks heralds a transformative era in network slicing, facilitating tailored service delivery and optimal resource utilization. Despite its promise, network slice optimization heavily relies on Deep Reinforcement Learning (DRL) models, often criticized for their black-box decision-making processes. this paper introduces a novel Composable eXplainable Reinforcement Learning (XRL) framework customized for distributed systems like 6G Network Slicing. the proposed framework leverages Large Language Models (LLMs) and Prompt Engineering techniques to elucidate DRL algorithms’ decision-making mechanisms, with a specific emphasis on user profiles. the latter transforms the inherently opaque nature of DRL into an interpretable textual format accessible not only to eXplainable AI (XAI) experts but also to diverse network slice provider stakeholders, engineers, leaders, and beyond. Experimental results underscore the efficacy of the proposed Composable XRL framework, showcasing substantial improvements in transparency and comprehensibility of DRL decisions within the context of 6G network slicing.
Single-camera distance measurement can provide a low-cost and setup-free experience. Several techniques to estimate the distance of the object of interest from a single image require calibration for specific parameter...
详细信息
Single-camera distance measurement can provide a low-cost and setup-free experience. Several techniques to estimate the distance of the object of interest from a single image require calibration for specific parameters and training for a model when machine learning approaches are implemented. this study introduces the use of a single calibration-free image to generate and evaluate a geometric model with a tilted-oriented object. ArUco fiducial marker is utilized since it can be detected easily using its pre-existing detection software. the perimeter, sum of diagonal lengths, and area of the pattern corners are the features for distance estimation. Our algorithm can determine the distance using only a single image in the training stage, which is considered well-balanced between complexity and accuracy in diverse low-cost, setup-free, or calibration-free applications.
this proceedings contains 12 papers. the proceedings of the VLDB Endowment (PVLDB) provides a high-quality publication service to the data management research community. this conference issue covers the topics which i...
Combining data from different sources into an integrated view is a recent trend taking advantage of the Internet of things (IoT) evolution over the last years. the fusion of different modalities has applications in va...
详细信息
ISBN:
(纸本)9783030983581;9783030983574
Combining data from different sources into an integrated view is a recent trend taking advantage of the Internet of things (IoT) evolution over the last years. the fusion of different modalities has applications in various fields, including healthcare and security systems. Human activity recognition (HAR) is among the most common applications of a healthcare or eldercare system. Inertial measurement unit (IMU) wearable sensors, like accelerometers and gyroscopes, are often utilized for HAR applications. In this paper, we investigate the performance of wearable IMU sensors along with vital signs sensors for HAR. A massive feature extraction, including both time and frequency domain features and transitional features for the vital signs, along with a feature selection method were performed. the classification algorithms and different early and late fusion methods were applied to a public dataset. Experimental results revealed that both IMU and vital signs achieve reasonable HAR accuracy and Fl-score among all the classes. Feature selection significantly reduced the number of features from both IMU and vital signs features while also improved the classification accuracy. the rest of the early and late level fusion methods also performed better than each modality alone, reaching an accuracy level of up to 95.32%.
暂无评论