Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702700
V. P. Bhardwaj
Parallel computation gives strength to the applications where high end and reliable computation is required. It is now aligned with the latest technologies of computer science. Blockchain technology, IoT, DevOps, AI and ML are some of the examples of latest technologies. At the core of almost every technology, it is observed that parallel computation is required and interconnection network gives strength to parallel computing applications. Therefore, author has shown his interest in this particular domain. In the present paper, author has compared two interconnection networks i.e. advanced omega network and irregular advance omega network. The comparison is founded on certain performance factors like bandwidth, and throughput. Both of these networks are coming into the categories of multistage interconnection networks, which have a huge contribution in the field of parallel computing. Further, result section shows that advanced omega network is better in terms of performance than the irregular advance omega network at network size 16, and it can be used as an alternative network for the applications where high performance computing is required.
{"title":"On Comparative Analysis of Advanced Omega Network and Irregular Advance Omega Network","authors":"V. P. Bhardwaj","doi":"10.1109/ICIIP53038.2021.9702700","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702700","url":null,"abstract":"Parallel computation gives strength to the applications where high end and reliable computation is required. It is now aligned with the latest technologies of computer science. Blockchain technology, IoT, DevOps, AI and ML are some of the examples of latest technologies. At the core of almost every technology, it is observed that parallel computation is required and interconnection network gives strength to parallel computing applications. Therefore, author has shown his interest in this particular domain. In the present paper, author has compared two interconnection networks i.e. advanced omega network and irregular advance omega network. The comparison is founded on certain performance factors like bandwidth, and throughput. Both of these networks are coming into the categories of multistage interconnection networks, which have a huge contribution in the field of parallel computing. Further, result section shows that advanced omega network is better in terms of performance than the irregular advance omega network at network size 16, and it can be used as an alternative network for the applications where high performance computing is required.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"260 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127788044","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}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702548
Prashanth Kannadaguli
Due to expanded praxis of social media, there is an elevated interest in the Natural Language Processing (NLP) of textual substance. Code swapping is a ubiquitous paradox in multilingual nation and the social communication shows mixing of a low resourced language with a highly resourced language mostly written in non-native script in the same text. It is essential to refine the code swapped text to support distinctive NLP tasks such as Machine Translation, Automated Conversational Systems and Sentiment Analysis (SA). The preeminent objective of SA is to identify and analyze the attitude, opinion, emotion or the sentiment in the dataset. Though there are multiple systems skilled on mono-dialectal dataset, all of them break down when it comes for code-diverse data because of the heightened intricacy of blending at various standards of text. Nonetheless, there exist a smaller number of assets for modelling such definitive code-mixed data and the Machine Learning or the Deep Learning algorithms enforcing supervised learning approach yield the better results compared to the unsupervised learning. Such datasets are available for Hindi-English, Tamil-English, Malayalam-English, Bengali-English, German-English, Spanish-English, Japanese-English, Arabic-English etc. Though our research is concentrated towards NLP for emotion and sentiment detection of Kannada, a vibrant south Indian language, to start with, we build the first ever platinum standard corpus for NLP applications of code-diverse text in Kannada-English, as there is no such resource in our native language. The performance analysis of our dataset through Krippendorff’s Alpha value of 0.89 indicates that it is a benchmark in development of Automatic Sentiment Analysis system for Kannada.
{"title":"A Code-Diverse Kannada-English Dataset For NLP Based Sentiment Analysis Applications","authors":"Prashanth Kannadaguli","doi":"10.1109/ICIIP53038.2021.9702548","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702548","url":null,"abstract":"Due to expanded praxis of social media, there is an elevated interest in the Natural Language Processing (NLP) of textual substance. Code swapping is a ubiquitous paradox in multilingual nation and the social communication shows mixing of a low resourced language with a highly resourced language mostly written in non-native script in the same text. It is essential to refine the code swapped text to support distinctive NLP tasks such as Machine Translation, Automated Conversational Systems and Sentiment Analysis (SA). The preeminent objective of SA is to identify and analyze the attitude, opinion, emotion or the sentiment in the dataset. Though there are multiple systems skilled on mono-dialectal dataset, all of them break down when it comes for code-diverse data because of the heightened intricacy of blending at various standards of text. Nonetheless, there exist a smaller number of assets for modelling such definitive code-mixed data and the Machine Learning or the Deep Learning algorithms enforcing supervised learning approach yield the better results compared to the unsupervised learning. Such datasets are available for Hindi-English, Tamil-English, Malayalam-English, Bengali-English, German-English, Spanish-English, Japanese-English, Arabic-English etc. Though our research is concentrated towards NLP for emotion and sentiment detection of Kannada, a vibrant south Indian language, to start with, we build the first ever platinum standard corpus for NLP applications of code-diverse text in Kannada-English, as there is no such resource in our native language. The performance analysis of our dataset through Krippendorff’s Alpha value of 0.89 indicates that it is a benchmark in development of Automatic Sentiment Analysis system for Kannada.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131344548","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}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702571
Shahid A. Malik, S. A. Parah, B. A. Malik
During its acquisition phase an ECG signal gets adulterated with distinct variants of undesirable noise thereby degrading its qualitative nature thereby inflicting a restraint on its clinical applicability. Hence it becomes imperative to design efficient methods to remove these artifacts specifically without deteriorating the signal quality. From classical approaches to modern digital methods, a multitude of methods have been reported in the literature for this purpose. In this paper, we have employed a computer-based hybrid approach that scrutinizes the denoising potential of VMD method. It proceeds by disintegrating an ECG signal polluted with high frequency PLI and low frequency noise into a band of VMFs with PLI distributed over lower order modes while as the low frequency noise distributed over the higher order modes. The higher order modes are then separately fed to an SWT system while as the sum of the lower order modes is fed to a non-local mean filter. Finally, the signal is reconstructed from the processed modes to generate a pure ECG signal free from artefacts. The prowess of the given method has been experimentally validated through the improvements in the three empirical parameters viz.: output SNR, cross-correlation coefficient and percentage root-mean-square difference. These parameters ascertain that the ECG signal has been efficiently denoised and faithfully reconstructed whilst maintaining and preserving its overall features. The experiments have been performed on the various recordings available online at MIT-BIH arrhythmia database.
{"title":"A VMD-SWT based ECG denoising technique","authors":"Shahid A. Malik, S. A. Parah, B. A. Malik","doi":"10.1109/ICIIP53038.2021.9702571","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702571","url":null,"abstract":"During its acquisition phase an ECG signal gets adulterated with distinct variants of undesirable noise thereby degrading its qualitative nature thereby inflicting a restraint on its clinical applicability. Hence it becomes imperative to design efficient methods to remove these artifacts specifically without deteriorating the signal quality. From classical approaches to modern digital methods, a multitude of methods have been reported in the literature for this purpose. In this paper, we have employed a computer-based hybrid approach that scrutinizes the denoising potential of VMD method. It proceeds by disintegrating an ECG signal polluted with high frequency PLI and low frequency noise into a band of VMFs with PLI distributed over lower order modes while as the low frequency noise distributed over the higher order modes. The higher order modes are then separately fed to an SWT system while as the sum of the lower order modes is fed to a non-local mean filter. Finally, the signal is reconstructed from the processed modes to generate a pure ECG signal free from artefacts. The prowess of the given method has been experimentally validated through the improvements in the three empirical parameters viz.: output SNR, cross-correlation coefficient and percentage root-mean-square difference. These parameters ascertain that the ECG signal has been efficiently denoised and faithfully reconstructed whilst maintaining and preserving its overall features. The experiments have been performed on the various recordings available online at MIT-BIH arrhythmia database.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"93 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132416097","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}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702587
G. Elizabeth Rani, G. Narasimha Murthy, Madhurapantula Abhiram, Harini Mohan, Tara Singh Naik, M. Sakthimohan
The recent Airlines management is facing lots of challenges and the pandemic has made it more critical. The airlines' industry needs to come up with a strong solution to uplift the airlines' sector and sophisticate the customers. In this paper, the main objective of the Airlines reservation system is to implement software using java that accompanies blockchain technology considering the airline sector challenges. It helps users to reserve tickets for air service and track the updated status periodically. Blockchain technology keeps the data secured and centralized providing efficient usage via mobile apps or online. The system provides an efficient user interface for both customers and stakeholders and analyzes the behavior of the customer and provides efficient results. This article also explains the demand price prediction and related challenges to be solved efficiently. All the above factors are considered and an efficient solution of application system using Java.
{"title":"An Automated Airlines Reservation Prediction System Using BlockChain Technology","authors":"G. Elizabeth Rani, G. Narasimha Murthy, Madhurapantula Abhiram, Harini Mohan, Tara Singh Naik, M. Sakthimohan","doi":"10.1109/ICIIP53038.2021.9702587","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702587","url":null,"abstract":"The recent Airlines management is facing lots of challenges and the pandemic has made it more critical. The airlines' industry needs to come up with a strong solution to uplift the airlines' sector and sophisticate the customers. In this paper, the main objective of the Airlines reservation system is to implement software using java that accompanies blockchain technology considering the airline sector challenges. It helps users to reserve tickets for air service and track the updated status periodically. Blockchain technology keeps the data secured and centralized providing efficient usage via mobile apps or online. The system provides an efficient user interface for both customers and stakeholders and analyzes the behavior of the customer and provides efficient results. This article also explains the demand price prediction and related challenges to be solved efficiently. All the above factors are considered and an efficient solution of application system using Java.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130520855","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}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702667
Shreya Kalta, Ravindara Bhatt
Heart disease is one of the diseases that are becoming a major cause of mortality throughout the world. A large population in the world is suffering from this problem. Considering the death rate and people suffering from heart diseases, reveals the early diagnosis of heart disease. The health care industry generates terabytes of data every day, which requires proper analysis and prediction of data which can be accomplished through data mining which acts as an intelligent diagnostic tool in heart disease diagnosis. In this research work two data mining classification algorithms are used which are Decision tree and Back-propagation network and are built using Python programming language on Anaconda’s Jupyter Notebook. The main purpose of this research is to identify and compare the best classification algorithm with the highest degree of accuracy, which will aid professionals in making decisions and diagnosing the probability of occurrence of heart disease in a patient. Thus preventing the loss of lives at the earliest. The heart disease dataset was obtained from Kaggle with 303 patient records and 14 essential clinical features and the output classifies whether or not a person has heart disease. After the comparative analysis the results proved that Back-propagation gives better results and shows greater accuracy which is 93% as compared to Decision tree.
{"title":"A Comparison Analysis of Heart Disease Dataset Using Decision Tree and Back-Propagation Network","authors":"Shreya Kalta, Ravindara Bhatt","doi":"10.1109/ICIIP53038.2021.9702667","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702667","url":null,"abstract":"Heart disease is one of the diseases that are becoming a major cause of mortality throughout the world. A large population in the world is suffering from this problem. Considering the death rate and people suffering from heart diseases, reveals the early diagnosis of heart disease. The health care industry generates terabytes of data every day, which requires proper analysis and prediction of data which can be accomplished through data mining which acts as an intelligent diagnostic tool in heart disease diagnosis. In this research work two data mining classification algorithms are used which are Decision tree and Back-propagation network and are built using Python programming language on Anaconda’s Jupyter Notebook. The main purpose of this research is to identify and compare the best classification algorithm with the highest degree of accuracy, which will aid professionals in making decisions and diagnosing the probability of occurrence of heart disease in a patient. Thus preventing the loss of lives at the earliest. The heart disease dataset was obtained from Kaggle with 303 patient records and 14 essential clinical features and the output classifies whether or not a person has heart disease. After the comparative analysis the results proved that Back-propagation gives better results and shows greater accuracy which is 93% as compared to Decision tree.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"98 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132665136","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}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702600
Shubhangi Chaturvedi, P. Khanna, A. Ojha
Smoke is the first sign of ignition of fire because smoke becomes visible when the fire starts. At this stage, fire can be effectively controlled by locating the smoke at the earliest. Smoke causes several health issues such as skin allergies and breathing problems in humans and animals. One of the biggest smoke emission sources is the industrial smoke. For environmental safety, various harmful gases emitting from industrial chimneys need to be monitored constantly. Further, increasing incidents of wildfire have also resulted in severe environmental degradation in recent years. Thus, detection of smoke and finding its location at early stage can help in mitigating fire hazards. Several vision based techniques have been proposed by researchers using traditional image processing techniques in the past to identify and segment smoke in images. In recent years, deep learning techniques have shown promising performance in smoke detection. In this paper, we present a comparative analysis of traditional image processing and recent deep learning based smoke segmentation techniques with focus on industrial and wildfire smoke.
{"title":"Comparative Analysis of Traditional and Deep Learning Techniques for Industrial and Wildfire Smoke Segmentation","authors":"Shubhangi Chaturvedi, P. Khanna, A. Ojha","doi":"10.1109/ICIIP53038.2021.9702600","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702600","url":null,"abstract":"Smoke is the first sign of ignition of fire because smoke becomes visible when the fire starts. At this stage, fire can be effectively controlled by locating the smoke at the earliest. Smoke causes several health issues such as skin allergies and breathing problems in humans and animals. One of the biggest smoke emission sources is the industrial smoke. For environmental safety, various harmful gases emitting from industrial chimneys need to be monitored constantly. Further, increasing incidents of wildfire have also resulted in severe environmental degradation in recent years. Thus, detection of smoke and finding its location at early stage can help in mitigating fire hazards. Several vision based techniques have been proposed by researchers using traditional image processing techniques in the past to identify and segment smoke in images. In recent years, deep learning techniques have shown promising performance in smoke detection. In this paper, we present a comparative analysis of traditional image processing and recent deep learning based smoke segmentation techniques with focus on industrial and wildfire smoke.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130160246","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}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702611
Nihashree Sarma, A. K. Talukdar, K. K. Sarma
Sign Language is a language which helps deaf and mute people for communication with hearing people. The aim of Indian Sign Language recognition (ISLR) is to understand the meaning of signs of speech impaired or hearing impaired person in the Indian region to interact with the society. This paper proposes for ISLR system in real-time based on the YOLOv3 Model and used in conjunction with Darknet-53 convolutional neural network. The system has been tested in real-time with 16 different signs for images and 7 signs for videos. The proposed model was labeled the sign language datasets in YOLO format. The sign language images are captured by the webcam for static sign language recognition and videos are recorded for dynamic sign language recognition. We achieved the accuracies for static and dynamic signs as 95.7% and 93.1%, respectively. The experimental results show that the proposed system can recognize both static and dynamic ISL signs effectively in real time with reduced computational time.
手语是一种帮助聋哑人与听力正常的人交流的语言。印度手语识别(Indian Sign Language recognition, ISLR)的目的是为了让印度地区的语言障碍或听力障碍人士了解手语的含义，以便与社会进行互动。本文提出了一种基于YOLOv3模型并结合Darknet-53卷积神经网络的实时ISLR系统。该系统已经在16种不同的图像标记和7种视频标记上进行了实时测试。该模型以YOLO格式对手语数据集进行标记。通过网络摄像头捕获静态手语图像进行识别，录制视频进行动态手语识别。我们实现了静态和动态标志的准确率分别为95.7%和93.1%。实验结果表明，该系统可以有效地实时识别静态和动态ISL符号，减少了计算时间。
{"title":"Real-Time Indian Sign Language Recognition System using YOLOv3 Model","authors":"Nihashree Sarma, A. K. Talukdar, K. K. Sarma","doi":"10.1109/ICIIP53038.2021.9702611","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702611","url":null,"abstract":"Sign Language is a language which helps deaf and mute people for communication with hearing people. The aim of Indian Sign Language recognition (ISLR) is to understand the meaning of signs of speech impaired or hearing impaired person in the Indian region to interact with the society. This paper proposes for ISLR system in real-time based on the YOLOv3 Model and used in conjunction with Darknet-53 convolutional neural network. The system has been tested in real-time with 16 different signs for images and 7 signs for videos. The proposed model was labeled the sign language datasets in YOLO format. The sign language images are captured by the webcam for static sign language recognition and videos are recorded for dynamic sign language recognition. We achieved the accuracies for static and dynamic signs as 95.7% and 93.1%, respectively. The experimental results show that the proposed system can recognize both static and dynamic ISL signs effectively in real time with reduced computational time.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129841702","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}
Pub Date : 2021-11-26DOI: 10.1109/iciip53038.2021.9702578
{"title":"Technical Programme Committee Members/Reviewers","authors":"","doi":"10.1109/iciip53038.2021.9702578","DOIUrl":"https://doi.org/10.1109/iciip53038.2021.9702578","url":null,"abstract":"","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134608550","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}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702677
L. R, A. Padyana
Diabetic Retinopathy (DR) is one of the complications of diabetes that impacts blood vessels of a retina because of increased blood sugar. So, it’s better to detect and treat at the initial stage. The biggest challenges are inadequate technology assistance for ophthalmologists and difficulty in the manual identification process. These issues can be addressed by technological advancement in the field of Artificial Intelligence for automizing the identification and detection process. An automatic detection helps to identify different stages of DR and helps ophthalmologists to provide treatment according to the stages in order to avoid vision loss. In this paper, proposed system aims to detect the various stages of DR that allows ophthalmologists to identify the DR at its different stage. The proposed system classifies the image data into defined classes using YOLO-RF. The proposed system compared with various traditional machine learning classifiers such as SVM, Decision Tree (DT), Random Forest (RF) and DL model such as YOLO. We have used data from the retinal fundus images of KAGGLE and IDRID. The result showed that proposed system YOLO-RF model performed with good accuracy of 99.3%, precision score of 97.2 and Recall of 99.1.
糖尿病视网膜病变(DR)是糖尿病的并发症之一，由于血糖升高而影响视网膜血管。因此，最好在早期发现和治疗。最大的挑战是对眼科医生的技术援助不足，以及人工识别过程的困难。这些问题可以通过人工智能领域的技术进步来解决，实现识别和检测过程的自动化。自动检测有助于识别DR的不同阶段，并帮助眼科医生根据不同阶段提供治疗，以避免视力丧失。本文提出的系统旨在检测DR的各个阶段，使眼科医生能够识别DR的不同阶段。该系统使用YOLO-RF对图像数据进行分类。该系统与各种传统的机器学习分类器(如SVM、Decision Tree (DT)、Random Forest (RF)和DL模型(如YOLO)进行了比较。我们使用的数据来自于KAGGLE和IDRID的视网膜眼底图像。结果表明，所提出的系统YOLO-RF模型准确率为99.3%，精密度评分为97.2，召回率为99.1。
{"title":"Detection of Diabetic Retinopathy in Retinal Fundus Image Using YOLO-RF Model","authors":"L. R, A. Padyana","doi":"10.1109/ICIIP53038.2021.9702677","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702677","url":null,"abstract":"Diabetic Retinopathy (DR) is one of the complications of diabetes that impacts blood vessels of a retina because of increased blood sugar. So, it’s better to detect and treat at the initial stage. The biggest challenges are inadequate technology assistance for ophthalmologists and difficulty in the manual identification process. These issues can be addressed by technological advancement in the field of Artificial Intelligence for automizing the identification and detection process. An automatic detection helps to identify different stages of DR and helps ophthalmologists to provide treatment according to the stages in order to avoid vision loss. In this paper, proposed system aims to detect the various stages of DR that allows ophthalmologists to identify the DR at its different stage. The proposed system classifies the image data into defined classes using YOLO-RF. The proposed system compared with various traditional machine learning classifiers such as SVM, Decision Tree (DT), Random Forest (RF) and DL model such as YOLO. We have used data from the retinal fundus images of KAGGLE and IDRID. The result showed that proposed system YOLO-RF model performed with good accuracy of 99.3%, precision score of 97.2 and Recall of 99.1.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130167790","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}
In the recent era, image tampering has become one of the threatening security problems in digital platforms. There are many software’s available for tampering with an image that depicts as an original image. Different tampering techniques are used to hide important portions from an image or document, one very common practice is a copy-move forgery that is quite impossible to distinguish with an open eye. Authentications of such images are an ardent research area in image processing and computer vision but still a challenging problem. This paper presents a method to identify image tampering that is based on SIFT (Scale Invariant Feature Transform) algorithm. SIFT descriptor is used to extract keypoint features from the input image and a hierarchical clustering algorithm is used to improve the accuracy of identifying the tampered location. The execution time of our proposed method is proportional to image resolution. If one portion of the image is copied and pasted on multiple locations on the same image, our proposed method can identify such occurrences. Finally, Homography is used to show the tampering points and their matching.
{"title":"Identification of Tampering Image Using SIFT Descriptor","authors":"Debjani Chakraborty, Sandeep Choudhury, Sanjib Kumar Dutta, Biswajit Haldar","doi":"10.1109/ICIIP53038.2021.9702601","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702601","url":null,"abstract":"In the recent era, image tampering has become one of the threatening security problems in digital platforms. There are many software’s available for tampering with an image that depicts as an original image. Different tampering techniques are used to hide important portions from an image or document, one very common practice is a copy-move forgery that is quite impossible to distinguish with an open eye. Authentications of such images are an ardent research area in image processing and computer vision but still a challenging problem. This paper presents a method to identify image tampering that is based on SIFT (Scale Invariant Feature Transform) algorithm. SIFT descriptor is used to extract keypoint features from the input image and a hierarchical clustering algorithm is used to improve the accuracy of identifying the tampered location. The execution time of our proposed method is proportional to image resolution. If one portion of the image is copied and pasted on multiple locations on the same image, our proposed method can identify such occurrences. Finally, Homography is used to show the tampering points and their matching.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130770233","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：