2021 Digital Image Computing: Techniques and Applications (DICTA)最新文献_第2页

Point Cloud Registration with Self-supervised Feature Learning and Beam Search 基于自监督特征学习和光束搜索的点云配准

2021 Digital Image Computing: Techniques and Applications (DICTA) Pub Date : 2021-11-01 DOI: 10.1109/DICTA52665.2021.9647267

Guofeng Mei

引用次数: 5

2021 Digital Image Computing: Techniques and Applications (DICTA) Pub Date : 2021-11-01 DOI: 10.1109/DICTA52665.2021.9647357

Priyanka Gupta, Sathvik Bhaskarpandit, Manik Gupta

{"title":"Similarity Learning based Few Shot Learning for ECG Time Series Classification","authors":"Priyanka Gupta, Sathvik Bhaskarpandit, Manik Gupta","doi":"10.1109/DICTA52665.2021.9647357","DOIUrl":"https://doi.org/10.1109/DICTA52665.2021.9647357","url":null,"abstract":"Using deep learning models to classify time series data generated from the Internet of Things (IoT) devices requires a large amount of labeled data. However, due to constrained resources available in IoT devices, it is often difficult to accommodate training using large data sets. This paper proposes and demonstrates a Similarity Learning-based Few Shot Learning for ECG arrhythmia classification using Siamese Convolutional Neural Networks. Few shot learning resolves the data scarcity issue by identifying novel classes from very few labeled examples. Few Shot Learning relies first on pretraining the model on a related relatively large database, and then the learning is used for further adaptation towards few examples available per class.Our experiments evaluate the performance accuracy with respect to K (number of instances per class) for ECG time series data classification. The accuracy with 5- shot learning is 92.25% which marginally improves with further increase in K. We also compare the performance of our method against other well-established similarity learning techniques such as Dynamic Time Warping (DTW), Euclidean Distance (ED), and a deep learning model - Long Short Term Memory Fully Convolutional Network (LSTM-FCN) with the same amount of data and conclude that our method outperforms them for a limited dataset size. For K=5, the accuracies obtained are 57%, 54%, 33%, and 92% approximately for ED, DTW, LSTM-FCN, and SCNN, respectively.","PeriodicalId":424950,"journal":{"name":"2021 Digital Image Computing: Techniques and Applications (DICTA)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131188043","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 3

GAN-based Spatial Transformation Adversarial Method for Disease Classification on CXR Photographs by Smartphones 基于gan的智能手机CXR照片疾病分类空间变换对抗方法

2021 Digital Image Computing: Techniques and Applications (DICTA) Pub Date : 2021-11-01 DOI: 10.1109/DICTA52665.2021.9647192

Chak Fong Chong, Xu Yang, Wei Ke, Yapeng Wang

{"title":"GAN-based Spatial Transformation Adversarial Method for Disease Classification on CXR Photographs by Smartphones","authors":"Chak Fong Chong, Xu Yang, Wei Ke, Yapeng Wang","doi":"10.1109/DICTA52665.2021.9647192","DOIUrl":"https://doi.org/10.1109/DICTA52665.2021.9647192","url":null,"abstract":"Deep learning has been successfully applied on Chest X-ray (CXR) images for disease classification. To support remote medical services (e.g., online diagnosis services), such systems can be deployed on smartphones by patients or doctors to take CXR photographs using the cameras on smartphones. However, photograph introduces visual artifacts such as blur, noises, light reflection, perspective transformation, moiré pattern, etc. plus unwanted background. Therefore, the classification accuracy of well-trained CNN models performed on the CXR photographs experiences drop significantly. Such challenge has not been solved properly in the literature. In this paper, we have compared various traditional image preprocessing methods on CXR photographs, including spatial transformation, background hiding, and various filtering methods. The combination of these methods can almost eliminate the negative impact of visual artifacts on the evaluation of 3 different single CNN models (Xception, DenseNet-121, Inception-v3), only 0.0018 AUC drop observed. However, such methods need user manually process the CXR photographs, which is inconvenient. Therefore, we have proposed a novel Generative Adversarial Network-based spatial transformation adversarial method (GAN-STAM) which can automatically transform the CXR region to the center and enlarge the CXR region in each CXR photograph, the classification accuracy has been significantly improved on CXR photographs from 0.8009 to 0.8653.","PeriodicalId":424950,"journal":{"name":"2021 Digital Image Computing: Techniques and Applications (DICTA)","volume":"117 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133720648","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 2

Overlapping Cell Nuclei Segmentation in Digital Histology Images using Intensity-based Contours 基于强度轮廓的数字组织学图像重叠细胞核分割

2021 Digital Image Computing: Techniques and Applications (DICTA) Pub Date : 2021-11-01 DOI: 10.1109/DICTA52665.2021.9647395

Md. Shamim Hossain, L. Armstrong, Jumana Abu-Khalaf, David M. Cook, P. Zaenker

{"title":"Overlapping Cell Nuclei Segmentation in Digital Histology Images using Intensity-based Contours","authors":"Md. Shamim Hossain, L. Armstrong, Jumana Abu-Khalaf, David M. Cook, P. Zaenker","doi":"10.1109/DICTA52665.2021.9647395","DOIUrl":"https://doi.org/10.1109/DICTA52665.2021.9647395","url":null,"abstract":"Automated nuclei segmentation techniques in histopathological image analysis continue to improve. The machine learning model requires the annotation of large data sets which is a time-consuming, expensive, and laborious process. This segmentation is also limited in detecting touching or overlapping nuclei and considers any overlapping nuclei as a single nucleus. This is due to low contrast images, occultation, and diversity of cell nuclei. This work proposes an automated overlapping nuclei segmentation model with a U-net and an intensity-based contour technique in order to address these issues. In a previous study, a U-net segmentation model was trained with synthetic data, which was generated using a GAN model, where a small number of histopathology data was used to generate the synthetic data. This reduced the data limitation and need for nuclei annotation in the deep learning model. Initially in this study, the overlapping nuclei regions were not considered for segmentation by the network. Hence, an intensity-based contour line is proposed to separate overlapping nuclei regions. The distance transformation is utilized to define the center of each nucleus. The identification of local minima followed by intensity-based gradient weights is applied to obtain the final segmentation line of overlapping nuclei. The boundary of the overlapping nuclei is refined, and noise is removed in order to clearly describe each nuclei region. The proposed method results in 91.6% accuracy in separating the overlapping nuclei compared to other existing methods.","PeriodicalId":424950,"journal":{"name":"2021 Digital Image Computing: Techniques and Applications (DICTA)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114867341","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

SCMNet: Shared Context Mining Network for Real-time Semantic Segmentation 实时语义分割的共享上下文挖掘网络

2021 Digital Image Computing: Techniques and Applications (DICTA) Pub Date : 2021-11-01 DOI: 10.1109/DICTA52665.2021.9647401

Tanmay Singha, Moritz Bergemann, Duc-Son Pham, A. Krishna

{"title":"SCMNet: Shared Context Mining Network for Real-time Semantic Segmentation","authors":"Tanmay Singha, Moritz Bergemann, Duc-Son Pham, A. Krishna","doi":"10.1109/DICTA52665.2021.9647401","DOIUrl":"https://doi.org/10.1109/DICTA52665.2021.9647401","url":null,"abstract":"Different architectures have been adopted for realtime scene segmentation. A popular design is the multi-branch approach in which multiple independent branches are deployed at the encoder side to filter input images at different resolutions. The main purpose is to reduce the computational cost and handle high resolution. However, independent branches do not contribute in the learning process. To address this issue, we introduce a novel approach in which two branches at the encoder share their knowledge whilst generating the global feature map. At each sharing point, the shared features will go through a new effective feature scaling module, called the Context Mining Module (CMM), which will refine the shared knowledge before passing it to the next stage. Finally, we introduce a new multidirectional feature fusion module which fuses deep contextual features with shallow features successively with accurate object localization. Our novel scene parsing model, termed SCMNet, produces 66.5% validation mIoU on the Cityscapes dataset and 78.6% on the Camvid dataset whilst having only 1.2 million parameters. Furthermore, the proposed model can efficiently handle higher resolution input images whilst having low computational cost. Our proposed model produces state-of-the-art results on Camvid.","PeriodicalId":424950,"journal":{"name":"2021 Digital Image Computing: Techniques and Applications (DICTA)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128479390","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 6

Use of uncertainty quantification as a surrogate for layer segmentation error in Stargardt disease retinal OCT images 利用不确定度量化替代Stargardt病视网膜OCT图像的层分割误差

2021 Digital Image Computing: Techniques and Applications (DICTA) Pub Date : 2021-11-01 DOI: 10.1109/DICTA52665.2021.9647154

D. Alonso-Caneiro, J. Kugelman, Janelle Tong, M. Kalloniatis, F. Chen, Scott A. Read, M. Collins

{"title":"Use of uncertainty quantification as a surrogate for layer segmentation error in Stargardt disease retinal OCT images","authors":"D. Alonso-Caneiro, J. Kugelman, Janelle Tong, M. Kalloniatis, F. Chen, Scott A. Read, M. Collins","doi":"10.1109/DICTA52665.2021.9647154","DOIUrl":"https://doi.org/10.1109/DICTA52665.2021.9647154","url":null,"abstract":"Semantic segmentation methods based on deep learning techniques have transformed the analysis of many medical imaging modalities, including the extraction of retinal layers from ocular optical coherence tomography images. Despite the high accuracy of these methods, the automatic techniques are not free of labelling errors, which means that a clinician may need to engage in the time-consuming process of reviewing the outcome of the segmentation method. Given this shortcoming, having access to segmentation techniques that can provide a confidence metric associated with the output (probability class map) are desirable. In this study, the use of Monte-Carlo dropout combined with a residual U-net architecture is explored as a way to provide segmentation pixel-wise prediction maps as well as corresponding uncertainty maps. While assessing the proposed network on a dataset of subjects with a retinal pathology (Stargardt disease), the uncertainty map exhibited a high correlation with the boundary error metric. Thus, confirming the potential of the technique to extract metrics that are a surrogate of the segmentation error. While the Monte-Carlo dropout seems to have no detrimental effect on performance, the uncertainty metric derived from this technique has potential for a range of important clinical (i.e. ranking of scans to be reviewed by a human expert) and research (i.e. network fine-tuning with a focus on high uncertainty/high error regions) applications.","PeriodicalId":424950,"journal":{"name":"2021 Digital Image Computing: Techniques and Applications (DICTA)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127185454","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

IoT-based Plant Health Analysis using Optical Sensors in Precision Agriculture 精准农业中基于物联网的光学传感器植物健康分析

2021 Digital Image Computing: Techniques and Applications (DICTA) Pub Date : 2021-11-01 DOI: 10.1109/DICTA52665.2021.9647066

H. Bagha, Ali Yavari, Dimitrios Georgakopoulos

{"title":"IoT-based Plant Health Analysis using Optical Sensors in Precision Agriculture","authors":"H. Bagha, Ali Yavari, Dimitrios Georgakopoulos","doi":"10.1109/DICTA52665.2021.9647066","DOIUrl":"https://doi.org/10.1109/DICTA52665.2021.9647066","url":null,"abstract":"To support the current population growth, modern agriculture must increase food production while reducing the use of water and other resources required for crop cultivation. Precision agriculture (PA) aims to achieve these via a variety of methods that include site-specific plant selection, variable rate irrigation and fertilisation, as well as site-specific pesticide and herbicide application. To determine the plant performance and health that drive such precision PA practices, PA solutions currently collect and analyse data from cameras and multispectral sensors. Technological advancements in the Internet of Things (IoT) and in the development of Unmanned Aerial Vehicles (UAV) in recent years have provided potential solutions for automating image acquisition and analysis that can advance such PA practices. This paper proposes 1) devising plant models from RGB and multi-spectral data, 2) using such models to guide the above PA practices. More specifically, the paper explores monitoring plants at different health and life cycle stages from fully green to completely dry and capturing related RGB and multi-spectral data in a controlled environment. These data are then analysed to create a model for each plant variety, which we refer to as the plant profile, that captures the combined colour and light reflectance of the plant over its life cycle and related health stages. The paper proposes using such plant variety profiles to determine the performance and health of the plants across entire crops. Finally, the paper discusses how UAVs and IoT can be used to automatically capture and analyse the images and multi-spectral data for advancing PA.","PeriodicalId":424950,"journal":{"name":"2021 Digital Image Computing: Techniques and Applications (DICTA)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130720664","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 2

ODAR: A Lightweight Object Detection Framework for Autonomous Driving Robots 面向自动驾驶机器人的轻量级目标检测框架

2021 Digital Image Computing: Techniques and Applications (DICTA) Pub Date : 2021-11-01 DOI: 10.1109/DICTA52665.2021.9647256

Le Hoang Duong, T. T. Huynh, Minh Tam Pham, Gwangzeen Ko, Jung Ick Moon, Jun Jo, N. Q. Hung

{"title":"ODAR: A Lightweight Object Detection Framework for Autonomous Driving Robots","authors":"Le Hoang Duong, T. T. Huynh, Minh Tam Pham, Gwangzeen Ko, Jung Ick Moon, Jun Jo, N. Q. Hung","doi":"10.1109/DICTA52665.2021.9647256","DOIUrl":"https://doi.org/10.1109/DICTA52665.2021.9647256","url":null,"abstract":"Object detection is an emerging and essential problem in recent years, which has been widely applied in many aspects of daily life such as video surveillance, self-driving robots, and automatic payment. The rapid development of deep learning models allows object detectors to work in real-time with high accuracy. However, such a sophisticated model often requires robust computing infrastructure such as powerful graphics processing units (GPUs). This requirement might cause a severe issue for embedded systems with small, power-efficient artificial intelligence (AI) systems like Jetson Nano, which are often restricted in both memory storage and computing sheer power. In this work, we aim to address this challenge by proposing a lightweight object detection framework that is specialized for the Internet of Things (IoT) devices with low-power processors such as Jetson Nano. In order to detect the object with different size, our framework employs a backbone residual CNN-based network as the feature extractor. We then design a multi-layer model to combine the feature at different levels of granularity, before using the processed feature to locate and classify the object. We also apply augmentation techniques to enhance the robustness of the framework to adversarial factors. Extensive experiments on real devices in many scenarios, such as autonomous cars or wireless robot recharging systems, showed that our technique can achieve nearly on par results with the state-of-the-art YOLOv5 while requires only one-fourth of computation power.","PeriodicalId":424950,"journal":{"name":"2021 Digital Image Computing: Techniques and Applications (DICTA)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126506919","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

Rapid Segmentation of Thoracic Organs using U-net Architecture 基于U-net结构的胸部器官快速分割

2021 Digital Image Computing: Techniques and Applications (DICTA) Pub Date : 2021-11-01 DOI: 10.1109/DICTA52665.2021.9647312

Hassan Mahmood, S. Islam, J. Hill, G. Tay

{"title":"Rapid Segmentation of Thoracic Organs using U-net Architecture","authors":"Hassan Mahmood, S. Islam, J. Hill, G. Tay","doi":"10.1109/DICTA52665.2021.9647312","DOIUrl":"https://doi.org/10.1109/DICTA52665.2021.9647312","url":null,"abstract":"Medical imaging provides a non-invasive method to diagnose, monitor, and plan the treatment of disease inside the human body. The increasing prevalence of radiological scanners and prescription of their use has presented a significant challenge for radiologists in accurately diagnosing disease whilst dealing with a growing number of scans to review. Recent advances in Artificial Intelligence (AI), especially in machine learning, are enabling researchers to improve the patient experience, enhance the planning of medical treatments and increase the rate of examination of scans. In this study,a 2-dimensional (2D) U-net based deep learning model was used to automatically segment five organs of interest from Computed Tomography (CT) scans of the thoracic region. Comparable results were achieved in comparison to the top seven models from a prior thoracic organ segmentation challenge. The framework can perform the segmentation tasks within 20 seconds, reducing workload for radiologists and increasing throughput. This study shows that a simple U-net based framework can be sufficient for the task at hand rather than pursuing much more complicated architectures, depending upon the complexity of the problem. Furthermore, we investigated the effect of 3D interpolation on dice scores in anticipation of further research applications in mapping segments to a 3D volume render. We find performance degradation with respect to the dice score after mapping the masks to original dimensions.","PeriodicalId":424950,"journal":{"name":"2021 Digital Image Computing: Techniques and Applications (DICTA)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115960835","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 1

HEp-2 Specimen Cell Detection and Classification Using Very Deep Convolutional Neural Networks-Based Cell Shape 基于深度卷积神经网络的HEp-2细胞形态检测与分类

2021 Digital Image Computing: Techniques and Applications (DICTA) Pub Date : 2021-11-01 DOI: 10.1109/DICTA52665.2021.9647184

Brandon Jorgensen, Khamael Al-Dulaimi, Jasmine Banks

{"title":"HEp-2 Specimen Cell Detection and Classification Using Very Deep Convolutional Neural Networks-Based Cell Shape","authors":"Brandon Jorgensen, Khamael Al-Dulaimi, Jasmine Banks","doi":"10.1109/DICTA52665.2021.9647184","DOIUrl":"https://doi.org/10.1109/DICTA52665.2021.9647184","url":null,"abstract":"The accurate detection and classification of HEp-2 specimen staining plays a key role in autoimmune disease diagnosis and transplantation assessment. Such detection and classification is challenging due to the abundant presence of highly overlapped cells, variations in cell densities, the variety of staining patterns, large numbers of cells per image, large data volumes and overfitting of features. In this paper, a robust technique is proposed to deal with images of all staining patterns and address these challenges. Very deep convolutional neural networks with a layer structure inspired by the standard architecture of the VGG-16 image is proposed for classification of HEp-2 staining cells based on cell shape and adapted to consider overfitting. Level set method using geometric active contours with morphological opening and Delaunay triangulation is used for cell segmentation and splitting. The cell segmentation method also considers overlapped cells. The proposed method has been tested and compared with other methods using Task-2 training dataset from competitions held on the ICPR2014 and ICPR2016. A extensive study demonstrates that the proposed method outperforms all other methods and promises to support the diagnosis of autoimmune diseases and allograft rejection prediction in future pathology practice, except for one method in this study which is slightly better than our proposed method.","PeriodicalId":424950,"journal":{"name":"2021 Digital Image Computing: Techniques and Applications (DICTA)","volume":"174 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133712332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 3