The determination of the signal's angle of arrival (AOA) is crucial for civilian and military applications. Popular AOA determination techniques such as MUSIC are often calculation intensive and require a priori k...
详细信息
The determination of the signal's angle of arrival (AOA) is crucial for civilian and military applications. Popular AOA determination techniques such as MUSIC are often calculation intensive and require a priori knowledge about signal frequency and incoming signals need to have the same frequency. In this paper, an efficient joint frequency-AOA determination method based on a 2-dimensional fast Fourier transform (2D-FFT) is proposed and investigated. The major advantages of the proposed method lie in its simplicity and the availability of efficient hardware/software designed for FFT calculation. Simulation results demonstrate the validity of the proposed method and its advantage over MUSIC in terms of calculation efficiency and better noise immunity. The results show that the proposed method outperforms MUSIC in a low signal-to-noise ratio (SNR) environment (SNR < - 4 dB) and that it requires CPU time merely a fraction of what is required by MUSIC. Therefore, the proposed method provides a feasible approach for realizing AOA detection in a real-timeprocessing device.
To monitor the condition of cupping spots in real-time during the operation of the automatic cupping machine, reduce the influence of the surrounding environment on the image, and improve the segmentation accuracy of ...
详细信息
To monitor the condition of cupping spots in real-time during the operation of the automatic cupping machine, reduce the influence of the surrounding environment on the image, and improve the segmentation accuracy of the cupping spots, this paper proposes a network called MCA-Deeplabv3+. Firstly, backbone network replaced by Mobilenetv2 to reduce the model size and improve feature extraction speed;Secondly, to further enhance the network's feature extraction capabilities, we added dilated convolution channels and integrated the CA attention mechanism into the ASPP module;Finally, data augmentation and brightness adjustment are performed on the dataset to improve the generalization of the model in different environments. The experimental results show that, in comparison with other segmentation models, MCA-Deeplabv3+performs the best in cupping spot segmentation, with mIoU and mPA reaching 93.90% and 96.73%, respectively. The practicality and effectiveness of the cupping spot segmentation model presented in this paper are thoroughly demonstrated.
High Efficiency video Coding (HEVC) and Multi-access Edge Computing (MEC) technologies can make real-time streaming media services available to users with reasonable bandwidth, but the computational complexity of HEVC...
详细信息
ISBN:
(纸本)9781728198354
High Efficiency video Coding (HEVC) and Multi-access Edge Computing (MEC) technologies can make real-time streaming media services available to users with reasonable bandwidth, but the computational complexity of HEVC tends to lead to increased energy consumption in these schemes. In this paper, we investigate the energy saving opportunities of utilizing a field-programmable gate array (FPGA) based HEVC encoder in edge media servers and devices. In practice, we analyze the energy impact of migrating our Kvazaar software HEVC intra encoder to Intel Arria 10 PCIe FPGA(s) on two platforms: 1) Nokia Airframe Cloud Server with 2.4 GHz dual 14-core Intel Xeon processors and 2) an embedded Jetson AGX Orin board with 2.2 GHz 12-core ARM processor. According to our experiments, FPGA encoding on these two platforms saved 76% and 86% of the energy taken up by software only encoding on Airframe, respectively. These results indicate the potential of FPGA-based video encoder acceleration in future green MEC architectures.
real-time online video super-resolution (VSR) on resource limited applications is a very challenging problem due to the constraints on complexity, latency and memory footprint, etc. Recently, a series of fast online V...
详细信息
ISBN:
(纸本)9781728198354
real-time online video super-resolution (VSR) on resource limited applications is a very challenging problem due to the constraints on complexity, latency and memory footprint, etc. Recently, a series of fast online VSR methods have been proposed to tackle this issue. In particular, attention based methods have achieved much progress by adaptively aligning or aggregating the information in preceding frames. However, these methods are still limited in network design to effectively and efficiently propagate the useful features in temporal domain. In this work, we propose a new fast online VSR algorithm with a flow-guided deformable attention propagation module, which leverages corresponding priors provided by a fast optical flow network in deformable attention computation and consequently helps propagating recurrent state information effectively and efficiently. The proposed algorithm achieves state-of-the-art results on widely-used benchmarking VSR datasets in terms of effectiveness and efficiency. Code can be found at https://***/IanYeung/FastOnlineVSR.
We propose an approach to the creation of a panorama viewport and objects detection within it in real-time on the base of the set of videos from the assembly of cameras. The task of the panorama viewport generation is...
详细信息
This paper presents the 6th edition of the "Drone-vs-Bird" detection challenge, jointly organized with the WOSDETC workshop within the IEEE International conference on Acoustics, Speech and Signal processing...
详细信息
This paper presents the 6th edition of the "Drone-vs-Bird" detection challenge, jointly organized with the WOSDETC workshop within the IEEE International conference on Acoustics, Speech and Signal processing (ICASSP) 2023. The main objective of the challenge is to advance the current state-of-the-art in detecting the presence of one or more Unmanned Aerial Vehicles (UAVs) in realvideo scenes, while facing challenging conditions such as moving cameras, disturbing environmental factors, and the presence of birds flying in the foreground. For this purpose, a video dataset was provided for training the proposed solutions, and a separate test dataset was released a few days before the challenge deadline to assess their performance. The dataset has continually expanded over consecutive installments of the Drone-vs-Bird challenge and remains openly available to the research community, for non-commercial purposes. The challenge attracted novel signal processing solutions, mainly based on deep learning algorithms. The paper illustrates the results achieved by the teams that successfully participated in the 2023 challenge, offering a concise overview of the state-of-the-art in the field of drone detection using video signal processing. Additionally, the paper provides valuable insights into potential directions for future research, building upon the main pros and limitations of the solutions presented by the participating teams.
Lossy compression reduces the data amount in the video by sacrificing quality, which leads to severe distortion, especially when videos are overly compressed. Con-sequently, many restoration methods have been proposed...
详细信息
Structural damages, such as structural looseness and structural cracks, are commonly observed as the root causes of failures in industrial plants. These issues have been extensively studied, and deep diagnostic tools ...
详细信息
Structural damages, such as structural looseness and structural cracks, are commonly observed as the root causes of failures in industrial plants. These issues have been extensively studied, and deep diagnostic tools have shown promise in identifying and addressing them. However, these tools rely on large amounts of data, which leads to computational burdens and time consumption. To tackle this challenge, a groundbreaking technique is proposed within the context of this study. The key innovation of this approach lies in its ability to integrate information from various processing functions and utilize an efficient feature bank that facilitates the execution of an effective feature learning method based on a multisource strategy. This novel research also focuses on the selection of transferable features from multiple distributions for diagnostics involving unseen failure distributions. By minimizing the mean squared error function, which is based on various source domains, the accuracy of diagnostics is significantly improved. Furthermore, the joint minimization of diagnostics independence concerning failure distribution, as well as the dimension of the transferable feature space between the source domains, leads to enhanced diagnostics speed and feature visualization. To validate the effectiveness of this approach, a real case study of a structural/machinery vibration dataset is conducted to address the multi-fault diagnosis problem, encompassing machinery health conditions, foundation looseness, and cracks under various operational conditions. The results obtained from this study demonstrate that the proposed algorithm performs remarkably well in real diagnostics scenarios involving unseen failure distributions.
Parking management systems play a crucial role in addressing parking shortages and operational challenges;however, high initial costs and infrastructure requirements often hinder their implementation. Edge computing o...
详细信息
Parking management systems play a crucial role in addressing parking shortages and operational challenges;however, high initial costs and infrastructure requirements often hinder their implementation. Edge computing offers a promising solution by reducing latency and network traffic, thus optimizing operational costs. Nonetheless, the limited computational resources of edge devices remain a significant challenge. This study developed a real-time vehicle occupancy detection system utilizing SSD-MobileNetv2 on edge devices to process video streams from multiple IP cameras. The system incorporates a dual-trigger mechanism, combining periodic triggers and parking space mask triggers, to optimize computational efficiency and resource usage while maintaining high accuracy and reliability. Experimental results demonstrated that the parking space mask trigger significantly reduced unnecessary AI model executions compared to periodic triggers, while the dual-trigger mechanism ensured consistent updates even under unstable network conditions. The SSD-MobileNetv2 model achieved a frame processingtime of 0.32 s and maintained robust detection performance with an F1-score of 0.9848 during a four-month field validation. These findings validate the suitability of the system for real-time parking management in resource-constrained environments. Thus, the proposed smart parking system offers an economical, viable, and practical solution that can significantly contribute to developing smart cities.
Due to poor illumination and low contrast, semantic segmentation of nighttimeimages faces major challenges. Various segmentation models with a large number of parameters are proposed to improve the performance but le...
详细信息
Due to poor illumination and low contrast, semantic segmentation of nighttimeimages faces major challenges. Various segmentation models with a large number of parameters are proposed to improve the performance but lead to an inability to process in realtime. To tackle these problems, we propose a real-time edge-guided bilateral network (EGBNet) for nighttime semantic segmentation. Considering the blurred details and low contrast of nighttimeimages, we propose a lightweight multi-dilation dense aggregation module and introduce an efficient edge head to improve the ability to distinguish target features from the nighttime background. Moreover, a self-adaptive feature fusion module is proposed for the bilateral segmentation network to enhance the feature representation and generalization ability by fully using multi-scale feature maps. To capture more useful information from limited nighttimeimages, we further use the knowledge distillation strategy to improve the segmentation performance. Extensive experiments on ACDC and BDD datasets demonstrate the effectiveness of our EGBNet by achieving a satisfactory trade-off between segmentation accuracy and inference speed. Specifically, EGBNet achieves 55.56% mIoU on the ACDC test set with 9.4 M parameters and 60FPS speed for a 1080 x 1920 input image on a single NVIDIA 2080Ti.
暂无评论