Anomaly detection in spacecraft telemetry is critical for the success and safety of space missions. Traditional methods often rely on forecasting and threshold techniques to identify anomalies [1]-[5]. This paper pres...
详细信息
ISBN:
(纸本)9798350384543;9798350384536
Anomaly detection in spacecraft telemetry is critical for the success and safety of space missions. Traditional methods often rely on forecasting and threshold techniques to identify anomalies [1]-[5]. This paper presents a comprehensive comparison of traditional forecast-based anomaly detection against two innovative classification methods, including a direct classification and an image classification through Gramian Angular Field (GAF) transforms [6], which have only been analysed in other domains but not for spacecraft anomaly detection. All our investigated systems leverage deeplearning architectures and use the popular real SMAP/MSL spacecraft data from [2]. Our findings suggest that direct classification provides a marginal but statistically significant improvement in anomaly detection over traditional methods. However, image classification, while less successful, offers promising directions for future research. The study aims to guide the selection of appropriate anomaly detection techniques for spacecraft telemetry and contribute to the advancement of automated monitoring systems in space missions.
This study presents a novel artificial intelligence model (AIM) for the real-time classification of 13 different road types in an autonomous vehicle. The model was developed based on a combination of a continuous wave...
详细信息
This study presents a novel artificial intelligence model (AIM) for the real-time classification of 13 different road types in an autonomous vehicle. The model was developed based on a combination of a continuous wavelet transform (CWT) and convolutional neural network (CNN). Previously, three methods have been used for the classification of road types, depending on the type of sensor. First, a camera sensor has been widely used because it can capture the road type directly. Second, a vibration sensor has been used, since the vibration level measured on the suspension or inside the tire depends on the road type. Finally, an acoustic sensor has been used, especially in measuring tire-pavement interaction noise (TPIN). In a previous feasibility study, an AIM was developed to classify road types using TPIN signals, which vary depending on the road type. It can distinguish between two road types: asphalt and snow. Recently, CNN has been widely used as an AIM for classification, but it is limited as the input size of the CNN should be optimized for real-timeprocessing due to its long calculation times, even for a 2D convolution process. Its input is image data, which can be produced through the CWT of the TPIN signal. This study proposes an AIM that can classify 13 different road surfaces in real-time while driving. In this study, a method to determine the optimal filter band and data length used for CWT is proposed. The method was developed based on the classification accuracy of an AIM. The developed AIM was successfully applied to the real-time classification of road types with an accuracy of 95% on a public road.
The fineness modulus(FM) represents the level of particle size of manufactured sand. real-time feedback of FM of manufactured sand is important for industrial sand production, but extracting the particle profile from ...
详细信息
The fineness modulus(FM) represents the level of particle size of manufactured sand. real-time feedback of FM of manufactured sand is important for industrial sand production, but extracting the particle profile from densely stacked images is a great challenge. In this study, a deeplearning and regression analysis -based online measurement method for FM of manufactured sand is proposed. Firstly, the real fineness modulus of the sand produced by the sand -making machine in realtime was obtained by the vibration -screening method(VSM). Then, the particle size fraction of larger particles (0.6-4.75 mm) was obtained based on machine vision combined with a convolutional neural network and imageprocessing. Secondly, a multiple linear regression model was developed for the percentage of particle size and FM. Finally, the percentage of particle size was substituted into the regression model as the independent variable to achieve a fast prediction of the unknown FM. The experimental results show that the maximum repeatability errors for FM of different manufactured sands are 0.09 and 0.13 respectively, and the maximum absolute errors of the FM prediction results are 0.18 and 0.17 respectively. The calculation efficiency and error level of this research method can meet the online testing at sand making sites.
Edible mushrooms are rich in nutrients;however,harvesting mainly relies on manual *** localization of each mushroom is necessary to enable a robotic arm to accurately pick edible *** studies used detection algorithms ...
详细信息
Edible mushrooms are rich in nutrients;however,harvesting mainly relies on manual *** localization of each mushroom is necessary to enable a robotic arm to accurately pick edible *** studies used detection algorithms that did not consider mushroom pixel-level *** these algorithms are combined with a depth map,the information is ***,in instance segmentation algorithms,convolutional neural network(CNN)-based methods are lightweight,and the extracted features are not *** guarantee real-time location detection and improve the accuracy of mushroom segmentation,this study proposed a new spatial-channel transformer network model based on Mask-CNN(SCT-Mask-RCNN).The fusion of Mask-RCNN with the self-attention mechanism extracts the global correlation outcomes of image features from the channel and spatial ***,Mask-RCNN was used to maintain a lightweight structure and extract local features using a spatial pooling pyramidal structure to achieve multiscale local feature fusion and improve detection *** results showed that the SCT-Mask-RCNN method achieved a segmentation accuracy of 0.750 on segm_Precision_mAP and detection accuracy of 0.638 on Bbox_Precision_*** to existing methods,the proposed method improved the accuracy of the evaluation metrics Bbox_Precision_mAP and segm_Precision_mAP by over 2%and 5%,respectively.
Video deepFakes are fake media created with deeplearning (DL) that manipulate a person's expression or identity. Most current deepFake detection methods analyze each frame independently, ignoring inconsistencies ...
详细信息
Due to the sparsity of point clouds obtained by LIDAR, the depth information is usually not complete and dense. The depth completion task is to recover dense depth information from sparse depth information. However, m...
详细信息
Due to the sparsity of point clouds obtained by LIDAR, the depth information is usually not complete and dense. The depth completion task is to recover dense depth information from sparse depth information. However, most of the current deep completion networks use RGB images as guidance, which are more like a processing method of information fusion. They are not valid when there is only sparse depth data and no other color information. Therefore, this paper proposes an information-reinforced completion network for a single sparse depth input. We use a multi-resolution dense progressive fusion structure to maximize the multi-scale information and optimize the global situation by point folding. At the same time, we re-aggregate the confidence and impose another depth constraint on the pixel depth to make the depth estimation closer to the ground trues. Our experimental results on KITTI and NYU Depth v2 datasets show that the proposed network achieves better results than other unguided deep completion methods. And it is excellent in both accuracy and real-time performance.
With the development of deeplearning, deep convolution neural networks for medical image segmentation tasks have become more and more complex in pursuit of higher accuracy. In most scenarios, medical image segmentati...
详细信息
ISBN:
(数字)9781665496209
ISBN:
(纸本)9781665496209
With the development of deeplearning, deep convolution neural networks for medical image segmentation tasks have become more and more complex in pursuit of higher accuracy. In most scenarios, medical image segmentation pursues accuracy rather than speed, However, real-time performance is crucial in some scenarios, such as surgical navigation and diagnosis of acute stroke. So design of high-precision, lightweight and real-time medical image segmentation network has become an urgent need. To this end, a novel lightweight dual-domain network (LDD-Net) has been proposed in this paper. LDD-Net is comprised of two branches, learning respectively from the frequency domain and the spatial domain. In the frequency domain branch, the image spatial resolution is compressed via discrete cosine transform to have a large receptive field, so that better semantic context features can be learned. In the spatial domain branch, high-resolution feature representations with more details are learned. Finally, the learned features of these two branches are fused to yield high accuracy with low computational cost. The proposed method has been validated on two medical image segmentation datasets to yield the state-of-the-art performances with greatly reduced inference time and parameters of the learned models.
Accurate identification of fish species is critical for applications in fisheries management., biodiversity monitoring., and conservation efforts. Conventional manual identification techniques take a lot of time and a...
详细信息
In this paper YOLOv8 deeplearning model is proposed for vehicle detection, classification, and counting for urban traffic surveillance applications on custom dataset. The model was trained with images and annotations...
详细信息
The contemporary landscape of sophisticated and intelligent jamming techniques presents significant challenges for traditional radar anti-jamming methods, necessitating advancements in radar operational capabilities. ...
详细信息
The contemporary landscape of sophisticated and intelligent jamming techniques presents significant challenges for traditional radar anti-jamming methods, necessitating advancements in radar operational capabilities. To address the intricate and dynamic nature of jamming scenarios, alongside the limitations in performance assurance of manually crafted anti-jamming strategies and the suboptimal real-time responsiveness of radar systems, this study introduces an intelligent decision-making model founded on deep reinforcement learning (DRL). This model is meticulously structured, comprising a defined action space, state space, and reward function. Concurrently, the paper advocates a novel radar anti-jamming strategy learning approach based on the deep Q-Network (DQN), adept at mitigating external malicious interference. This approach enhances the integration efficiency and the doppler frequency resolution in radar echo processing. Comparative simulation outcomes affirm the superiority of the proposed intelligent decision model and training methodology over established methods like Proximal Policy Optimization (PPO) and Q-learning. Notably, the model demonstrates enhanced jamming suppression, robust generalization capabilities, accelerated response times, and a significant augmentation in the radar's autonomous decision-making prowess.
暂无评论