Pub Date : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544904
Er.Dipra Mitra, Shikha Gupta, D. Srivastava
Human Eye is one of the most dedicated organs. Eyes help the human beings to see the world around them. But by using the power of computer vision and machine learning, researchers are working on the way to detect human eyes and find a way to control a computer system. The authors' main idea is to work on an algorithm that will enable human beings to access any computing platform just by their eyes. The main challenge is to use eye movements to track and access a computing platform. The foremost thing is that the algorithm will work on complex images or video feed regarding any constraints on the background or any color pigment of human skin complexion or tone.
{"title":"A computer vision-based Algorithmic approach towards Eye motion Access —A review","authors":"Er.Dipra Mitra, Shikha Gupta, D. Srivastava","doi":"10.1109/ICIRCA51532.2021.9544904","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544904","url":null,"abstract":"Human Eye is one of the most dedicated organs. Eyes help the human beings to see the world around them. But by using the power of computer vision and machine learning, researchers are working on the way to detect human eyes and find a way to control a computer system. The authors' main idea is to work on an algorithm that will enable human beings to access any computing platform just by their eyes. The main challenge is to use eye movements to track and access a computing platform. The foremost thing is that the algorithm will work on complex images or video feed regarding any constraints on the background or any color pigment of human skin complexion or tone.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129013314","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544795
M. Kavitha, P. Srinivas, P. Kalyampudi, Choragudi S. F, S. Srinivasulu
Anomaly detection is a vital research problem among the different domains intrusion detection, fraud detection, device health monitoring, fault data detection, event detection in sensor networks. Anomalies mean an outlier, noise, novelties, exceptions which do not match the expected behavior of the system. Machine learning techniques work well in identifying these abnormal patterns. In this paper, the unsupervised clustering technique K-means, and its variation K-medoids partitioning are applied to detect anomalies. Sensor-embedded wearable devices are allowing smart healthcare services for people even in remote areas. These devices support continuous monitoring of people's health and allow the caregivers to provide better health assistance. Early-stage anomaly detection in such types of smart healthcare practices increases the efficiency of health services. In experimental discussion, K-means, and K-medoids partitioning clustering algorithms are assessed, and their performance is addressed.
{"title":"Machine Learning Techniques for Anomaly Detection in Smart Healthcare","authors":"M. Kavitha, P. Srinivas, P. Kalyampudi, Choragudi S. F, S. Srinivasulu","doi":"10.1109/ICIRCA51532.2021.9544795","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544795","url":null,"abstract":"Anomaly detection is a vital research problem among the different domains intrusion detection, fraud detection, device health monitoring, fault data detection, event detection in sensor networks. Anomalies mean an outlier, noise, novelties, exceptions which do not match the expected behavior of the system. Machine learning techniques work well in identifying these abnormal patterns. In this paper, the unsupervised clustering technique K-means, and its variation K-medoids partitioning are applied to detect anomalies. Sensor-embedded wearable devices are allowing smart healthcare services for people even in remote areas. These devices support continuous monitoring of people's health and allow the caregivers to provide better health assistance. Early-stage anomaly detection in such types of smart healthcare practices increases the efficiency of health services. In experimental discussion, K-means, and K-medoids partitioning clustering algorithms are assessed, and their performance is addressed.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"101 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124560181","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544847
N. K, S. S., S. S
Heart Disease refers to a broad range of heart-related health problems. Heart disease is currently the world's most serious public health issue. Many organizations have made extensive use of data mining. Data mining in healthcare is becoming trendy, if not extremely important. The health sector nowadays produces a significant volume of complex data about individuals, diagnosis of diseases, clinical notes, medical equipment, and so on. The objective is to know about the various data mining methods that have evolved to forecast heart problems. According to the findings, a Random forest with 15 features outstripped all such data-mining methods.
{"title":"Prediction and Analysis of Heart disease using Data mining Algorithms","authors":"N. K, S. S., S. S","doi":"10.1109/ICIRCA51532.2021.9544847","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544847","url":null,"abstract":"Heart Disease refers to a broad range of heart-related health problems. Heart disease is currently the world's most serious public health issue. Many organizations have made extensive use of data mining. Data mining in healthcare is becoming trendy, if not extremely important. The health sector nowadays produces a significant volume of complex data about individuals, diagnosis of diseases, clinical notes, medical equipment, and so on. The objective is to know about the various data mining methods that have evolved to forecast heart problems. According to the findings, a Random forest with 15 features outstripped all such data-mining methods.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129469133","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544859
V. V, Denil C Verghese, Mohammed Arshu P T, Randheer Ramesh K, Subin T G
After the introduction of machine learning, it has gone through lots of research and development which resulted in an explosion of usage in many fields. Developing such a model is not an easy task and it requires extensive domain knowledge and skills. This paper presents Autofhm, a python library used for automated machine learning. This tool automates the steps followed for the machine learning model creation such as feature engineering, model selection, and hyperparameter optimization. For a given dataset, Autofhm generates new deeper features which could increase the performance of the model. Then it selects the best performing model along with the suitable hyperparameter combinations based on the feature engineered dataset. The Autofhm is tested on 5 classification tasks and 5 regression tasks and the results demonstrate that, Autofhm gives good results with lesser time when compared to state-of-the-art frameworks like TPOT.
{"title":"Autofhm: A Python Library for Automated Machine Learning","authors":"V. V, Denil C Verghese, Mohammed Arshu P T, Randheer Ramesh K, Subin T G","doi":"10.1109/ICIRCA51532.2021.9544859","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544859","url":null,"abstract":"After the introduction of machine learning, it has gone through lots of research and development which resulted in an explosion of usage in many fields. Developing such a model is not an easy task and it requires extensive domain knowledge and skills. This paper presents Autofhm, a python library used for automated machine learning. This tool automates the steps followed for the machine learning model creation such as feature engineering, model selection, and hyperparameter optimization. For a given dataset, Autofhm generates new deeper features which could increase the performance of the model. Then it selects the best performing model along with the suitable hyperparameter combinations based on the feature engineered dataset. The Autofhm is tested on 5 classification tasks and 5 regression tasks and the results demonstrate that, Autofhm gives good results with lesser time when compared to state-of-the-art frameworks like TPOT.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130647584","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544610
Suresh M B, Abhishek M R
The Kidney stones are a hard collection of salt and minerals, often calcium and uric acid that form in the kidneys. The majority of persons with kidney stones do not recognize them at first, and their organs gradually deteriorate. For surgical procedures, it is critical to determine the exact and precise location of a kidney stone. Speckle noise is present in most ultrasound images, which cannot be removed by humans. The paper consists of problems of kidney stones in the human body and detection mechanisms by using Image processing techniques. The Techniques like preprocessing, segmentation and Morphological Analysis. The Results of techniques are evaluated based on the output parameters and analyzed to conclude the methods working efficiently.
{"title":"Kidney Stone Detection Using Digital Image Processing Techniques","authors":"Suresh M B, Abhishek M R","doi":"10.1109/ICIRCA51532.2021.9544610","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544610","url":null,"abstract":"The Kidney stones are a hard collection of salt and minerals, often calcium and uric acid that form in the kidneys. The majority of persons with kidney stones do not recognize them at first, and their organs gradually deteriorate. For surgical procedures, it is critical to determine the exact and precise location of a kidney stone. Speckle noise is present in most ultrasound images, which cannot be removed by humans. The paper consists of problems of kidney stones in the human body and detection mechanisms by using Image processing techniques. The Techniques like preprocessing, segmentation and Morphological Analysis. The Results of techniques are evaluated based on the output parameters and analyzed to conclude the methods working efficiently.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"133 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123393171","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544790
M. Jagadeeswari, C. S. Manikandababu, R. Balaji, A. G. Kumar
Virtual learning platforms are important for the future of education, especially during unprecedented times like the current covid-19 pandemic. Such learning platforms are expected to be interactive and help students communicate better with teachers and other students even virtually. This research work intends to develop a virtual learning platform in the form of a website that allows teachers to connect with students via individual and group video conferencing, create basic quizzes for the students, easily evaluate the quizzes and monitor student attendance. This website would also be useful for the students as it allows them to learn better by understanding answers for the graded quizzes. It also allows the students to view their obtained marks, check their attendance, have one-to-one video interaction with the teachers using a WEBRTC technique and Python Django framework, and, much more, all in a single platform. All the data are stored and manipulated in the MYSQL database. Thereby serving as a one-stop approach for every need of a student without having multiple websites and thereby creating a hassle out of it. The entire front end was developed entirely using web technologies like HTML, CSS, Javascript.
{"title":"Virtual Learning Assistance for Students","authors":"M. Jagadeeswari, C. S. Manikandababu, R. Balaji, A. G. Kumar","doi":"10.1109/ICIRCA51532.2021.9544790","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544790","url":null,"abstract":"Virtual learning platforms are important for the future of education, especially during unprecedented times like the current covid-19 pandemic. Such learning platforms are expected to be interactive and help students communicate better with teachers and other students even virtually. This research work intends to develop a virtual learning platform in the form of a website that allows teachers to connect with students via individual and group video conferencing, create basic quizzes for the students, easily evaluate the quizzes and monitor student attendance. This website would also be useful for the students as it allows them to learn better by understanding answers for the graded quizzes. It also allows the students to view their obtained marks, check their attendance, have one-to-one video interaction with the teachers using a WEBRTC technique and Python Django framework, and, much more, all in a single platform. All the data are stored and manipulated in the MYSQL database. Thereby serving as a one-stop approach for every need of a student without having multiple websites and thereby creating a hassle out of it. The entire front end was developed entirely using web technologies like HTML, CSS, Javascript.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"34 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114092523","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544730
Vinuja S, K. A, Kaushek Kumar T R, U. R, K. R
Diabetic Retinopathy (DR), a complexity induced by high blood sugar level is found to degrade the light-sensitive tissue retina by harming the blood vessels present in the region. In this work, the two models of InceptionV3 and Xception have been used as a Diabetic Retinopathy classifier to classify the given images on a ranking from 0 to 4. The APTOS 2019 dataset containing colour fundus images of various levels of severity of DR have been used to train the two models. The two models are further evaluated based on four different combinations of data pre-processing and data augmentation techniques. The Gaussian blur method was utilized for the pre-processing of the dataset. Data augmentation methods like image rotation, horizontal and vertical flips and uniform brightening were used. After comparing the performance of the two models, it was found that the Xception gave the best performance with an accuracy of 93.10% when both preprocessing and augmentation were performed on the dataset. InceptionV3 yielded an accuracy of 91.90% after employing both pre-processing and augmentation on the dataset.
{"title":"Performance Analysis of Diabetic Retinopathy Classification using CNN","authors":"Vinuja S, K. A, Kaushek Kumar T R, U. R, K. R","doi":"10.1109/ICIRCA51532.2021.9544730","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544730","url":null,"abstract":"Diabetic Retinopathy (DR), a complexity induced by high blood sugar level is found to degrade the light-sensitive tissue retina by harming the blood vessels present in the region. In this work, the two models of InceptionV3 and Xception have been used as a Diabetic Retinopathy classifier to classify the given images on a ranking from 0 to 4. The APTOS 2019 dataset containing colour fundus images of various levels of severity of DR have been used to train the two models. The two models are further evaluated based on four different combinations of data pre-processing and data augmentation techniques. The Gaussian blur method was utilized for the pre-processing of the dataset. Data augmentation methods like image rotation, horizontal and vertical flips and uniform brightening were used. After comparing the performance of the two models, it was found that the Xception gave the best performance with an accuracy of 93.10% when both preprocessing and augmentation were performed on the dataset. InceptionV3 yielded an accuracy of 91.90% after employing both pre-processing and augmentation on the dataset.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116191455","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9545039
S. M, Namrata Kolkar, Suman G S, Keerti D Kulkarni
Communication with the outside world or the caretakers is one of the major challenges for the people with disabilities. In this work, the authors propose a realtime algorithm to detect morse code from the eye blinks in a series from a live camera. The proposed algorithm detects eye landmarks and estimates the level of eye-opening using eye aspect ratio (EAR). The low-cost software decodes morse code precisely based on the duration of the eyes closed or opened which in turn is translated to English Language. So, this system is intended to provide an alternative form of communication for people with disabilities and to convey confidential messages.
{"title":"Morse Code Detector and Decoder using Eye Blinks","authors":"S. M, Namrata Kolkar, Suman G S, Keerti D Kulkarni","doi":"10.1109/ICIRCA51532.2021.9545039","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9545039","url":null,"abstract":"Communication with the outside world or the caretakers is one of the major challenges for the people with disabilities. In this work, the authors propose a realtime algorithm to detect morse code from the eye blinks in a series from a live camera. The proposed algorithm detects eye landmarks and estimates the level of eye-opening using eye aspect ratio (EAR). The low-cost software decodes morse code precisely based on the duration of the eyes closed or opened which in turn is translated to English Language. So, this system is intended to provide an alternative form of communication for people with disabilities and to convey confidential messages.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"14 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113956946","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544726
G. Thilagavathi, G. Priyadharshini, A. M, Boopika A M, Swetha S V
Social media play a vital role in this information era. Twitter is one of the important microblogging platform where people can share information known to them. Often these tweets are about local events. News agencies report on local events, but the time taken for an agency to analyse, investigate and report on the event can be substantial. Twitter users share their views and information about a particular event by posting tweets. These tweets can be used to identify whether the event occurred or not. Event detection from twitter data has gained importance nowadays. Our proposed system analyses tweets from a given geographical region to determine if an event occurred. The system then report the most descriptive tweet associated with an event occurred in that particular region. By the proposed system, it would be a quick way to alert people about an event occurring in their locality. In this, we split data into clusters based on location, identifies the tweet which exceeds the threshold, and then group the tweets based on similarity. The clustering models DBSCAN and HDBSCAN are employed to eliminate noise from the data and cluster similar tweets. Our system converts each tweet into a vector and normalise using TF-IDF technique. Finally, tweets which are similar on the same event will be analysed and collected. People can be notified of local events occurring before news outlets can report them when it is implemented in real time. The application varies on the type of event detected using our system. The News stations can also be intimated about the event so that they can explore further.
{"title":"Detection of Social and Newsworthy events using Tweet Analysis","authors":"G. Thilagavathi, G. Priyadharshini, A. M, Boopika A M, Swetha S V","doi":"10.1109/ICIRCA51532.2021.9544726","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544726","url":null,"abstract":"Social media play a vital role in this information era. Twitter is one of the important microblogging platform where people can share information known to them. Often these tweets are about local events. News agencies report on local events, but the time taken for an agency to analyse, investigate and report on the event can be substantial. Twitter users share their views and information about a particular event by posting tweets. These tweets can be used to identify whether the event occurred or not. Event detection from twitter data has gained importance nowadays. Our proposed system analyses tweets from a given geographical region to determine if an event occurred. The system then report the most descriptive tweet associated with an event occurred in that particular region. By the proposed system, it would be a quick way to alert people about an event occurring in their locality. In this, we split data into clusters based on location, identifies the tweet which exceeds the threshold, and then group the tweets based on similarity. The clustering models DBSCAN and HDBSCAN are employed to eliminate noise from the data and cluster similar tweets. Our system converts each tweet into a vector and normalise using TF-IDF technique. Finally, tweets which are similar on the same event will be analysed and collected. People can be notified of local events occurring before news outlets can report them when it is implemented in real time. The application varies on the type of event detected using our system. The News stations can also be intimated about the event so that they can explore further.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"47 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114015869","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544765
Shekhar Karanwal
The influence of light changes makes the task of feature extraction more difficult for the local descriptors. Most of local descriptors sacrifices their performance in harsh lightning variations. Some uses pre-processing approach & some uses gradient based methods (with local descriptors) to improve accuracy. In this work, a novel Enhanced Local Descriptor (ELD) is introduced by taking advantages of two well behaved descriptors in harsh lightning changes. These 2 are Compound Local Binary Pattern (CLBP) & Median Robust Extended LBP based on Neighborhood Intensity (MRELBP-NI). CLBP is characterized by Sign & Magnitude details, and MRELBP-NI is characterized by Median & Mean statistics. Both of them are very essential in controlling harsh light variations. By merging features of both a discriminant descriptor ELD is gained. FLDA is taken further for size contraction & SVMs is used for matching. ELD achieves stupendous outcomes on Extended Yale B (EYB) dataset. ELD wholly outstrip the singly implemented descriptors & many methods from literature. ELD secure best accuracy of 93.42%. There is no pre-processing & the gradient based methods are used.
光照变化的影响使得局部描述子的特征提取任务更加困难。大多数局部描述符在恶劣的闪电变化中牺牲了它们的性能。有些使用预处理方法，有些使用基于梯度的方法(带有局部描述符)来提高精度。本文提出了一种新的增强局部描述子(Enhanced Local Descriptor, ELD)，利用两个描述子在强闪电变化中表现良好的优点。这两个是复合局部二值模式(CLBP)和基于邻域强度的中值鲁棒扩展LBP (MRELBP-NI)。CLBP以Sign & Magnitude细节表征，MRELBP-NI以Median & Mean统计特征表征。两者在控制强光变化方面都是非常重要的。通过合并两者的特征，得到一个判别描述符ELD。进一步采用FLDA进行尺寸收缩，采用svm进行匹配。ELD在扩展耶鲁B (EYB)数据集上取得了惊人的成果。ELD完全超越了文献中单个实现的描述符和许多方法。ELD的准确度为93.42%。没有预处理&使用基于梯度的方法。
{"title":"An Enhanced Local Descriptor (ELD) for Face Recognition","authors":"Shekhar Karanwal","doi":"10.1109/ICIRCA51532.2021.9544765","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544765","url":null,"abstract":"The influence of light changes makes the task of feature extraction more difficult for the local descriptors. Most of local descriptors sacrifices their performance in harsh lightning variations. Some uses pre-processing approach & some uses gradient based methods (with local descriptors) to improve accuracy. In this work, a novel Enhanced Local Descriptor (ELD) is introduced by taking advantages of two well behaved descriptors in harsh lightning changes. These 2 are Compound Local Binary Pattern (CLBP) & Median Robust Extended LBP based on Neighborhood Intensity (MRELBP-NI). CLBP is characterized by Sign & Magnitude details, and MRELBP-NI is characterized by Median & Mean statistics. Both of them are very essential in controlling harsh light variations. By merging features of both a discriminant descriptor ELD is gained. FLDA is taken further for size contraction & SVMs is used for matching. ELD achieves stupendous outcomes on Extended Yale B (EYB) dataset. ELD wholly outstrip the singly implemented descriptors & many methods from literature. ELD secure best accuracy of 93.42%. There is no pre-processing & the gradient based methods are used.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126429874","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：