2023
2023
-
Record 541 of
Title:Research on Modeling and Simulation of Full Link Noise in CCD Camera System
Author(s):Bu, Fan(1); Yao, Dalei(1); Yang, Yongqing(1); Cao, Weicheng(1)Source: Proceedings of SPIE - The International Society for Optical Engineering Volume: 12747 Issue: null Article Number: 127470M DOI: 10.1117/12.2689111 Published: 2023Abstract:Through in-depth research on various noise sources and characteristics of the full link of the CCD camera system, a mathematical model of the CCD camera system noise was established, and the mathematical model of the noise was simulated and analyzed using MATLAB digital simulation software. At the same time, indoor noise testing of the CCD camera was conducted, and the simulation results were basically consistent with the measured results, verifying the correctness of the noise mathematical model. These research conclusions lay a reliable theoretical foundation for the subsequent search for accurate CCD noise suppression methods. ? 2023 SPIE. All rights reserved.Accession Number: 20235115263408 -
Record 542 of
Title:A prediction model of microcirculation disorder in myocardium based on ultrasonic images
Author(s):Tian, Mingjun(1); Zheng, Minjuan(1); Qiu, Shi(2); Song, Yang(3)Source: Journal of Ambient Intelligence and Humanized Computing Volume: 14 Issue: 6 Article Number: null DOI: 10.1007/s12652-022-04440-5 Published: June 2023Abstract:The primary cause of coronary heart disease is abnormal myocardial circulatory perfusion. Microcirculation is the key link of myocardial oxygen supply and plays a major role in myocardial blood supply. We intended to employ myocardial load contrast enhanced ultrasound (MCSE) coupled with artificial intelligence to accurately evaluate the microcirculation of patients. (1) Based on the convolutional neural network, the framework of myocardial vessel extraction was constructed to extract myocardial vessels and accurately monitor the myocardium. (2) From the perspective of visual perception, the salient region algorithm was proposed to identify the perfusion signals according to the texture features and grayscale features, and the quantitative indexes of myocardial perfusion were obtained. (3) Combined with the imaging features and clinical features, the early diagnosis, efficacy evaluation, and risk stratification of coronary heart disease microcirculation disorders were evaluated. The results of this study lead to effectively evaluating and predicting myocardial microcirculation with AOM reaching 84% and the algorithm can aid medical professionals in diagnosis and therapy. ? 2022, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.Accession Number: 20224212900991 -
Record 543 of
Title:Longitudinal Structural MRI Data Prediction in Nondemented and Demented Older Adults via Generative Adversarial Convolutional Network
Author(s):Song, Liyao(1); Wang, Quan(2); Li, Haiwei(2); Fan, Jiancun(1); Hu, Bingliang(2)Source: Neural Processing Letters Volume: 55 Issue: 2 Article Number: null DOI: 10.1007/s11063-022-10922-6 Published: April 2023Abstract:Alzheimer’s disease (AD) is the most common cause of dementia and threatens the health of millions of people. Early stage diagnosis of AD is critical for improving clinical outcomes and longitudinal magnetic resonance imaging (MRI) data collection can be used to monitor the progress of each patient. However, missing data is a common problem in longitudinal AD studies. The main factors come from subject dropouts and failed scans. This hinders the acquisition of longitudinal sequences that consist of multi-time-point magnetic resonance (MR) images at relatively uniform intervals. In this paper, we present a generative adversarial convolutional network to predict missing structural MRI data. In particular, we include multiple MRI scans as a temporal sequence collected at different times and determine the spatio-temporal relationship between the different scans in the proposed network. We adopt residual bottlenecks in the generator to decrease parameter values and deepen the network. In order to make full use of the longitudinal information, our discriminator classifies not only real MR images from generated MR images, but also fake sequences from real sequences in which the longitudinal MR images for all time points come from the dataset, only the last MR image comes from the generator. Results of our experiment show that our method performs more accurately for the longitudinal structural MRI data prediction of a brain afflicted with AD. ? 2022, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.Accession Number: 20223212552154 -
Record 544 of
Title:Active Interactive Labelling Massive Samples for Object Detection
Author(s):Zhang, Jingwei(1); Zhang, Mingguang(1); Guo, Yi(2); Qiu, Mengyu(1)Source: Proceedings - SUI 2023: ACM Symposium on Spatial User Interaction Volume: null Issue: null Article Number: 34 DOI: 10.1145/3607822.3616407 Published: October 13, 2023Abstract:Aerial object detection is the process of detecting objects in remote sensing images, such as aerial or satellite imagery. However, due to the unique characteristics and challenges of remote sensing images, such as large image sizes and dense distribution of small objects, annotating the data is time-consuming and costly. Active learning methods can reduce the cost of labeling data and improve the model's generalization ability by selecting the most informative and representative unlabeled samples. In this paper, we studied how to apply active learning techniques to remote sensing object detection tasks and found that traditional active learning frameworks are not suitable. Therefore, we designed a remote sensing task-oriented active learning framework that can more efficiently select representative samples and improve the performance of remote sensing object detection tasks. In addition, we proposed an adaptive weighting loss to further improve the generalization ability of the model in unlabeled areas. A large number of experiments conducted on the remote sensing dataset DOTA-v2.0 showed that applying various classical active learning methods to the new active learning framework can achieve better performance. ? 2023 ACM.Accession Number: 20234615045966 -
Record 545 of
Title:Simulating Human Visual System Based on Vision Transformer
Author(s):Qiu, Mengyu(1); Guo, Yi(2); Zhang, Mingguang(1); Zhang, Jingwei(1); Lan, Tian(1); Liu, Zhilin(1)Source: Proceedings - SUI 2023: ACM Symposium on Spatial User Interaction Volume: null Issue: null Article Number: 35 DOI: 10.1145/3607822.3616408 Published: October 13, 2023Abstract:The human visual system (HVS) is capable of responding in real-time to complex visual environments. During the process of freely observing visual scenes, predicting eye movements and visual fixations is a task known as scanpath prediction, which aims to simulate the HVS. In this paper, we propose a visual transformer-based model to study the attentional processes of the human visual system in analyzing visual scenes, thereby achieving scanpath prediction. This technology has important applications in human-computer interaction, virtual reality, augmented reality, and other fields. We have significantly simplified the workflow of scanpath prediction and the overall model architecture, achieving performance superior to existing methods. ? 2023 ACM.Accession Number: 20234615045967 -
Record 546 of
Title:Design of multi-frequency point, high-isolation switches for micro-channel plate data acquisition
Author(s):Yang, Yihao(1,2); Gou, Yongsheng(1,2); Yang, Yang(1,2); Feng, Penghui(2); Wang, Bo(2); Liu, Baiyu(1,2); Wei, Jianan(1,2); Tian, Jinshou(1,2); Zhao, Wei(1,2)Source: Review of Scientific Instruments Volume: 94 Issue: 10 Article Number: 104713 DOI: 10.1063/5.0159975 Published: October 1, 2023Abstract:In order to replace the phosphor screen of a proximity-gated x-ray framing camera with a readout circuit using a time-interleaved structure, this paper carries out the design of a high-isolation RF switch. In this paper, a Metal-Oxide-Semiconductor Field Effect Tube (MOSFET) switching circuit is designed to achieve high isolation and low insertion loss at 0.5-3 GHz. This solves the problem that the switching circuit cannot be turned off properly due to the parasitic capacitance of MOSFETs in the process of RF signal transmission, resulting in signal feedthrough. It also ensures that the input signal can be transmitted to the output intact when the switching circuit is turned on. High isolation is achieved by using parallel resonance to increase the voltage division and series resonance to leak the current. The switch achieves 76 dB isolation and 0.07 dB insertion loss at 1 GHz frequency. Isolation is increased by adding parallel branches near the 2 and 3 GHz frequency points, achieving greater than 33 dB isolation from 0.5 to 3 GHz. ? 2023 Author(s).Accession Number: 20234515013537 -
Record 547 of
Title:Multi-Prior Graph Autoencoder with Ranking-Based Band Selection for Hyperspectral Anomaly Detection
Author(s):Wang, Nan(1,2); Shi, Yuetian(1,2); Li, Haiwei(1,3); Zhang, Geng(1,3); Li, Siyuan(1,3); Liu, Xuebin(1,3)Source: Remote Sensing Volume: 15 Issue: 18 Article Number: 4430 DOI: 10.3390/rs15184430 Published: September 2023Abstract:Hyperspectral anomaly detection (HAD) is an important technique used to identify objects with spectral irregularity that can contribute to object-based image analysis. Latterly, significant attention has been given to HAD methods based on Autoencoders (AE). Nevertheless, due to a lack of prior information, transferring of modeling capacity, and the "curse of dimensionality", AE-based detectors still have limited performance. To address the drawbacks, we propose a Multi-Prior Graph Autoencoder (MPGAE) with ranking-based band selection for HAD. There are three main components: the ranking-based band selection component, the adaptive salient weight component, and the graph autoencoder. First, the ranking-based band selection component removes redundant spectral channels by ranking the bands by employing piecewise-smooth first. Then, the adaptive salient weight component adjusts the reconstruction ability of the AE based on the salient prior, by calculating spectral-spatial features of the local context and the multivariate normal distribution of backgrounds. Finally, to preserve the geometric structure in the latent space, the graph autoencoder detects anomalies by obtaining reconstruction errors with a superpixel segmentation-based graph regularization. In particular, the loss function utilizes (Formula presented.) and adaptive salient weight to enhance the capacity of modeling anomaly patterns. Experimental results demonstrate that the proposed MPGAE effectively outperforms other state-of-the-art HAD detectors. ? 2023 by the authors.Accession Number: 20234014835265