The proceedings contain 9 papers. The topics discussed include: towards sustainable deployment of microservices over the Cloud-IoT continuum, with FREEDA;towards enabling observability of energy demand, with NEST;urge...
ISBN:
(纸本)9798400706417
The proceedings contain 9 papers. The topics discussed include: towards sustainable deployment of microservices over the Cloud-IoT continuum, with FREEDA;towards enabling observability of energy demand, with NEST;urgent edge computing;optimizing resource allocation in the edge: a minimum weighted vertex cover approach;towards energy-aware execution and offloading of serverless functions;reactive autoscaling of Kubernetes nodes;OUTFIT: crowdsourced data feeding noise maps in digital twins;encoding consistency: optimizing self-driving reliability with real-time speed data;and striking trade-off between high performance and energy efficiency in an edge computing application for detecting floating plastic debris.
The proceedings contain 3 papers. The topics discussed include: analyzing the energy consumption of synchronous and asynchronous checkpointing strategies;debugging MPI implementations via reduction-to-primitives;and e...
ISBN:
(纸本)9781665475723
The proceedings contain 3 papers. The topics discussed include: analyzing the energy consumption of synchronous and asynchronous checkpointing strategies;debugging MPI implementations via reduction-to-primitives;and emergency backup for scientific applications.
Autonomous mobile robot navigation is a challenging area of research due to various physical, hardware, and software issues. In this research, an autonomous robot system has been developed, which incorporates a visual...
详细信息
ISBN:
(纸本)9798350372977;9798350372984
Autonomous mobile robot navigation is a challenging area of research due to various physical, hardware, and software issues. In this research, an autonomous robot system has been developed, which incorporates a visual inertial navigation system (VINS) with the goal that the robot could conduct automated university campus tours. Mapping and state estimation rely on the accuracy acquired from fusing the data from the cameras and inertial measurement units (IMU). The fusion of these two sensors makes VINS systems more accurate and robust. We created a custom stereo inertial system and performed extensive evaluation to mitigate calibration issues such as noise and bias for accurate state estimation. The custom sensor infrastructure can be mounted on any mobile system. Furthermore, we tested our evaluation methodology on two challenging benchmark datasets, namely, EuRoC and TUM to determine the precision of the state estimator. Experiments show that VINS achieved highly accurate results in terms of calibration, reprojection, and trajectory estimation.
The upgrade and iteration of the artificial intelligence system has been enough to support it to cope with various complex tasks, such as intelligent analysis and intelligent tracking of images, which all require visu...
详细信息
Autonomous robot navigation relies on accurate motion estimation. Stereo visual Odometry is used to estimate the motion of a camera by analyzing the visual information captured by a stereo pair of cameras. visual odom...
详细信息
ISBN:
(纸本)9798350372977;9798350372984
Autonomous robot navigation relies on accurate motion estimation. Stereo visual Odometry is used to estimate the motion of a camera by analyzing the visual information captured by a stereo pair of cameras. visual odometry refers to the process of estimating the camera's movement by tracking the visual features in consecutive frames of a video sequence. In a stereo camera system, two cameras are positioned at known baseline separation. The images captured by the cameras are used to create a 3D representation of a scene. Stereo visual odometry leverages the disparity between corresponding points in the pair of stereo images to estimate the motion of camera. We propose a stereo visual odometry pipeline and perform a comprehensive comparison of different trajectory computation approaches on the challenging KITTI dataset. Different feature detection methods, feature matching, filter matching distance, stereo matching, different types of motion estimation, optimization techniques such as loop closure and bundle adjustment were tested in terms of error metric and computational time. SGBM matcher performed better than the BM matcher on depth and disparity estimation. SIFT was found to outperform the ORB feature detector. Brute Force feature matcher outperformed FLANN matcher. The lowest MAE error rate of 32.54 was obtained with SGBM matcher, brute force approach, PnP RANSAC, and filter matching distance of 0.3 on SIFT features.
UAV usually combines the Global Navigation Satellite System (GNNS) and the Inertial Navigation System (INS) to realize real-time localization and navigation. However, UAVs may lose GNSS signals due to complex terrain ...
详细信息
ISBN:
(纸本)9798400717048
UAV usually combines the Global Navigation Satellite System (GNNS) and the Inertial Navigation System (INS) to realize real-time localization and navigation. However, UAVs may lose GNSS signals due to complex terrain environments and electronic interference, resulting in localization errors beyond expectations. To solve the problem of UAV positioning under the condition of GNNS rejection, relevant scholars have proposed the absolute vision localization(AVL) technology, which avoids the cumulative error compared with the relative vision localization technology. However, the absolute vision localization algorithm is complex and time-consuming, and it cannot be deployed on low-computing power devices such as drones, which limits its application. This paper explores a visual localization scheme that can be deployed on embedded devices to meet computing power requirements and improve the speed of visual matching localization. We segment the satellite image, use the Kalman filter to predict the motion trajectory of the UAV, select the appropriate map for image matching, and run on the embedded device;the average time for a single localization is less than 200ms. Compared with other AVL methods, the matching speed is significantly improved, and the matching speed is not limited by the size of the satellite map, which verifies the feasibility of visual localization by UAV in practical applications.
In many fields, such as medical imaging, electronic games, and security monitoring, it is very necessary to perform target image recognition and processing on computers. Through image processing and visual communicati...
详细信息
In this paper, an edge computing box design method based on deep learning and embedded system is proposed to address the problems of complex harbor operating environment and frequent safety accidents. The computing bo...
详细信息
ISBN:
(纸本)9798400710025
In this paper, an edge computing box design method based on deep learning and embedded system is proposed to address the problems of complex harbor operating environment and frequent safety accidents. The computing box is based on Nvidia's Jetson Nano hardware architecture, and a Pedestrian net deep neural network model specialized in human and vehicle detection is developed through TensorRT and Deepstream. The computational box designed by this method supports secondary development and fast iteration, and can be flexibly adapted to various mainstream video processing AI models. Experiments show that the computational box has good computational real-time performance and extremely fast inference speed, which guarantees the deploy ability of port security and intelligent supervision in embedded terminal equipment.
Percepti-Alarm is a groundbreaking endeavor aimed at revolutionizing the lives of the visually impaired. It introduces a transformative solution by integrating a high-tech camera onto the conventional white cane, offe...
详细信息
The proceedings contain 193 papers. The topics discussed include: visual analysis of artificial intelligence in short video research based on CiteSpace;optimization design of financial shared services based on improve...
ISBN:
(纸本)9798331519032
The proceedings contain 193 papers. The topics discussed include: visual analysis of artificial intelligence in short video research based on CiteSpace;optimization design of financial shared services based on improved algorithms and artificial intelligence;research on the application of artificial intelligence technology in microelectronics testing and fault diagnosis;artificial intelligence method for v2g oriented power distribution network control strategy and capacity analysis;research on indoor lighting optimization control system based on artificial intelligence algorithm;a model for evaluating the effectiveness of news dissemination under the combination of big data and artificial intelligence;and mechanical and electronic engineering design and control technology under artificial intelligence.
暂无评论