Pub Date : 2021-09-24DOI: 10.1109/ICSES52305.2021.9633948
Swapna, Tangelapalli, P. Saradhi, R. Pandya, S. Iyer
The increasing demand for high-speed data services, such as mobile gaming, Augmented/Virtual Reality (AR/VR) applications, vehicular communications, Internet of Everything (IoE), and haptic internet, results in high user densification in 5G and beyond networks. Moreover, the ultra-dense user scenarios raise the challenge of increased interference due to the highly shared spatial resources and unknown Channel State Information (CSI). Therefore, the optimal channel estimation helps in interference cancellation; however, the conventional channel estimation techniques are imperfect. On the other hand, the Deep Learning (DL) approach confers the potential solution for the channel estimation. In this paper, we implement the Convolutional Neural Network (CNN) dL architecture for channel estimation over the range of values of SNR for Single Input Single Output OFDM network. The proposed DL-CNN approach demonstrates a 94.30% reduction in Mean Square Error (MSE) compared to the traditional interpolation method-based channel estimation at different values of SNR considering the dense scenario.
{"title":"Deep Learning Oriented Channel Estimation for Interference Reduction for 5G","authors":"Swapna, Tangelapalli, P. Saradhi, R. Pandya, S. Iyer","doi":"10.1109/ICSES52305.2021.9633948","DOIUrl":"https://doi.org/10.1109/ICSES52305.2021.9633948","url":null,"abstract":"The increasing demand for high-speed data services, such as mobile gaming, Augmented/Virtual Reality (AR/VR) applications, vehicular communications, Internet of Everything (IoE), and haptic internet, results in high user densification in 5G and beyond networks. Moreover, the ultra-dense user scenarios raise the challenge of increased interference due to the highly shared spatial resources and unknown Channel State Information (CSI). Therefore, the optimal channel estimation helps in interference cancellation; however, the conventional channel estimation techniques are imperfect. On the other hand, the Deep Learning (DL) approach confers the potential solution for the channel estimation. In this paper, we implement the Convolutional Neural Network (CNN) dL architecture for channel estimation over the range of values of SNR for Single Input Single Output OFDM network. The proposed DL-CNN approach demonstrates a 94.30% reduction in Mean Square Error (MSE) compared to the traditional interpolation method-based channel estimation at different values of SNR considering the dense scenario.","PeriodicalId":6777,"journal":{"name":"2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES)","volume":"15 1","pages":"1-4"},"PeriodicalIF":0.0,"publicationDate":"2021-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81639230","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 : 2021-09-24DOI: 10.1109/ICSES52305.2021.9633896
T. Sharma
This paper divulges an advanced dynamic model of comparator that contributes towards fast-rate of comparison, better slew-rate and lesser amount of power consumption. The preamplifier referred dynamic technique in comparators is extensively utilized in analogue to digital converters (ADCs). It provides positive feedback technique to rejuvenate the signal from analogue to full sway digital. When the output of preamplifier phase approaches to power supply, the added power is consumed by the overall circuit. In the proposed implementation of comparator model, the voltage swaying of preamplifier phase is restricted to one-half of the power supply. The new technique gives rise to lesser power and has high slew-rate. This paper depicts the comparison between separate dynamic comparator models. The simulation result is carried out using Pyxis tool (Mentor Graphics) in 180nm technology.
{"title":"Implementation of Power Efficient Dynamic Comparator at 180 nm Process Technology for high-speed applications","authors":"T. Sharma","doi":"10.1109/ICSES52305.2021.9633896","DOIUrl":"https://doi.org/10.1109/ICSES52305.2021.9633896","url":null,"abstract":"This paper divulges an advanced dynamic model of comparator that contributes towards fast-rate of comparison, better slew-rate and lesser amount of power consumption. The preamplifier referred dynamic technique in comparators is extensively utilized in analogue to digital converters (ADCs). It provides positive feedback technique to rejuvenate the signal from analogue to full sway digital. When the output of preamplifier phase approaches to power supply, the added power is consumed by the overall circuit. In the proposed implementation of comparator model, the voltage swaying of preamplifier phase is restricted to one-half of the power supply. The new technique gives rise to lesser power and has high slew-rate. This paper depicts the comparison between separate dynamic comparator models. The simulation result is carried out using Pyxis tool (Mentor Graphics) in 180nm technology.","PeriodicalId":6777,"journal":{"name":"2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES)","volume":"22 1","pages":"1-4"},"PeriodicalIF":0.0,"publicationDate":"2021-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81820211","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 : 2021-09-24DOI: 10.1109/ICSES52305.2021.9633972
B. Lavanya, C. Shanthi
Malware manual analysis still requires formula rules to verify that malicious samples are considered suspicious. Find the source of their software and malware as part of the code anatomy. To solve the security problem of the malware caused by the Android operating system, an efficient hybrid detection scheme is proposed for Android malware as the previous methods have not been efficient enough to detect advanced malware to limit/prevent damage. Machine learning technology provides the main novelty with high efficiency and low overhead. To verify that, this proposed gradient Conventional Recursive Neural Classifier (GCRNC) algorithm is feasible and many extensive malware data sets have been tested to prove its efficacy. The method has been classified into three stages: preprocessing, feature selection, and classification. The first preprocessing stage is based on Count Vectordistributionused to remove and extract the file types from the specified data set. Before classification, the feature is selected using the Adaboost Random Decision Tree Selection (ARDTS) method. The dataset uses are established to train first, and it is used with the expert weight assigned to each attribute by the domain expert. The rules are established based on the absolute rights assigned to this organization. The value of each selected feature is extracted and stored with the corresponding category label. The values are established based on the absolute rights assigned to this organization. A classification algorithm based on Gradient Conventional Recursive Neural Classifier (GCRNC) has been proposed to improve the achieved functional classification performance by only contributing to the effective classification process useful to classifying android malicious software datasets.
{"title":"Gradient Conventional Recursive Neural Classifier Algorithm to Analyze the Malicious Software Detection Using Machine Learning","authors":"B. Lavanya, C. Shanthi","doi":"10.1109/ICSES52305.2021.9633972","DOIUrl":"https://doi.org/10.1109/ICSES52305.2021.9633972","url":null,"abstract":"Malware manual analysis still requires formula rules to verify that malicious samples are considered suspicious. Find the source of their software and malware as part of the code anatomy. To solve the security problem of the malware caused by the Android operating system, an efficient hybrid detection scheme is proposed for Android malware as the previous methods have not been efficient enough to detect advanced malware to limit/prevent damage. Machine learning technology provides the main novelty with high efficiency and low overhead. To verify that, this proposed gradient Conventional Recursive Neural Classifier (GCRNC) algorithm is feasible and many extensive malware data sets have been tested to prove its efficacy. The method has been classified into three stages: preprocessing, feature selection, and classification. The first preprocessing stage is based on Count Vectordistributionused to remove and extract the file types from the specified data set. Before classification, the feature is selected using the Adaboost Random Decision Tree Selection (ARDTS) method. The dataset uses are established to train first, and it is used with the expert weight assigned to each attribute by the domain expert. The rules are established based on the absolute rights assigned to this organization. The value of each selected feature is extracted and stored with the corresponding category label. The values are established based on the absolute rights assigned to this organization. A classification algorithm based on Gradient Conventional Recursive Neural Classifier (GCRNC) has been proposed to improve the achieved functional classification performance by only contributing to the effective classification process useful to classifying android malicious software datasets.","PeriodicalId":6777,"journal":{"name":"2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES)","volume":"83 1","pages":"1-7"},"PeriodicalIF":0.0,"publicationDate":"2021-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90602979","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 : 2021-09-24DOI: 10.1109/ICSES52305.2021.9633877
M. Pushpavalli, P. Abirami, P. Sivagami, V. Geetha, M. Kavitha
This analyst for the most part centers around the steadiness examination of the multi-input DC converter framework. This MI contains two sources where wind and sun oriented can connect to the system. The performance of Multi-input DC Converters designed by the boost type with different converters configurations is Multi-input Boost, Buck-Boost, Sepic, Cuk, Zeta and KY Boost DC Converter. The Performance of the above converters analyzed in the closed-loop. Among this Multi-input KY boost, DC converter behaviour is superior to other MI DC Converters. MI KY boost DC converter is the most acceptable converter in terms of stability, time-domain analysis, and ripples. By comparing this, multi-input KY boost converter consumes less rise time, fall time and overshoot in percentage. Similarly, steady-state error (ESS) shows very low compared to other converters. Bode plot of a MI converter, measured gain margin in decibel and phase margin in degrees. Bode determines transfer function and pole-zero plot used to determine the balance of the plot. Dynamic conduct is tried and check with the assistance of MATLAB reproduction.
{"title":"Stability Analysis of Multi-Input DC-DC Converters","authors":"M. Pushpavalli, P. Abirami, P. Sivagami, V. Geetha, M. Kavitha","doi":"10.1109/ICSES52305.2021.9633877","DOIUrl":"https://doi.org/10.1109/ICSES52305.2021.9633877","url":null,"abstract":"This analyst for the most part centers around the steadiness examination of the multi-input DC converter framework. This MI contains two sources where wind and sun oriented can connect to the system. The performance of Multi-input DC Converters designed by the boost type with different converters configurations is Multi-input Boost, Buck-Boost, Sepic, Cuk, Zeta and KY Boost DC Converter. The Performance of the above converters analyzed in the closed-loop. Among this Multi-input KY boost, DC converter behaviour is superior to other MI DC Converters. MI KY boost DC converter is the most acceptable converter in terms of stability, time-domain analysis, and ripples. By comparing this, multi-input KY boost converter consumes less rise time, fall time and overshoot in percentage. Similarly, steady-state error (ESS) shows very low compared to other converters. Bode plot of a MI converter, measured gain margin in decibel and phase margin in degrees. Bode determines transfer function and pole-zero plot used to determine the balance of the plot. Dynamic conduct is tried and check with the assistance of MATLAB reproduction.","PeriodicalId":6777,"journal":{"name":"2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES)","volume":"41 1","pages":"1-7"},"PeriodicalIF":0.0,"publicationDate":"2021-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91517647","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 : 2021-09-24DOI: 10.1109/ICSES52305.2021.9633914
D. Indira, Kavitha Chaduvula
For computer users, data security is a major concern. In the medical field the data must be present in a hidden form rather than original. Each of us has essential medical data that we want to keep safe from others. Bit Plane Complexity Segmentation (BPCS) is described in this article as a method for hiding information within images in various planes. The algorithm can be used with the RSA public key encryption scheme to enhance security in data transmission. It is important to determine a threshold (approximate) value so that the stego and original images are the same, and so that the hiding capacity is greater with less computing cost. This process imposes a number of restrictions. This paper presents two modules encryption and decryption while transferring medical text in a hidden mode. Error rate is calculating with PSNR, BER and MER with three bit planes Red, Blue and Green.
{"title":"A Detailed Survey on Bit Plane Complexity Segmentation (BPCS) and RSA Algorithm for Secured Medical Data Transfer","authors":"D. Indira, Kavitha Chaduvula","doi":"10.1109/ICSES52305.2021.9633914","DOIUrl":"https://doi.org/10.1109/ICSES52305.2021.9633914","url":null,"abstract":"For computer users, data security is a major concern. In the medical field the data must be present in a hidden form rather than original. Each of us has essential medical data that we want to keep safe from others. Bit Plane Complexity Segmentation (BPCS) is described in this article as a method for hiding information within images in various planes. The algorithm can be used with the RSA public key encryption scheme to enhance security in data transmission. It is important to determine a threshold (approximate) value so that the stego and original images are the same, and so that the hiding capacity is greater with less computing cost. This process imposes a number of restrictions. This paper presents two modules encryption and decryption while transferring medical text in a hidden mode. Error rate is calculating with PSNR, BER and MER with three bit planes Red, Blue and Green.","PeriodicalId":6777,"journal":{"name":"2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES)","volume":"2 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2021-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84870129","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 : 2021-09-24DOI: 10.1109/ICSES52305.2021.9633903
Priyanka Gupta, A. P. Shukla
Lung Cancer is the world's fastest-growing cancer & is detected mainly at an early stage. Various modalities of medical imaging, such as computed tomography (CT) have been employed to reduce delays in diagnosis. So far, numerous machine learning architectures have been used by researchers to categorize lung nodules captured in CT scans into benign or cancerous. In this article, we proposed a novel 8-layer two-architecture of a three-dimensional deep convolutional neural network called AlexNet to classifying benign & malignant nodules from CT-Scan images. The Deep neural network extracts the features automatically. We apply binary cross-entropy to our proposed network's loss functionimprovetraining precision and validation accuracy of the model with 99% and 97% respectively.
{"title":"Improving Accuracy of Lung Nodule Classification Using AlexNet Model","authors":"Priyanka Gupta, A. P. Shukla","doi":"10.1109/ICSES52305.2021.9633903","DOIUrl":"https://doi.org/10.1109/ICSES52305.2021.9633903","url":null,"abstract":"Lung Cancer is the world's fastest-growing cancer & is detected mainly at an early stage. Various modalities of medical imaging, such as computed tomography (CT) have been employed to reduce delays in diagnosis. So far, numerous machine learning architectures have been used by researchers to categorize lung nodules captured in CT scans into benign or cancerous. In this article, we proposed a novel 8-layer two-architecture of a three-dimensional deep convolutional neural network called AlexNet to classifying benign & malignant nodules from CT-Scan images. The Deep neural network extracts the features automatically. We apply binary cross-entropy to our proposed network's loss functionimprovetraining precision and validation accuracy of the model with 99% and 97% respectively.","PeriodicalId":6777,"journal":{"name":"2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES)","volume":"30 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2021-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86532274","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 : 2021-09-24DOI: 10.1109/ICSES52305.2021.9633960
P. Malathi, S. Suganthidevi
Cloud based communication is a tremendous growth in the world and which provides the services through virtualized resources with support of internet. In cloud computing, optimized storage techniques require to store the hugea mount of data. Performing the data deduplication technique over the encrypted data consider as the major challenging task. Eliminating the redundant data file from the optimized storage, which results the minimizing the band width and reduces the cost and disk usage. In Existing approach, the most of the research work are using Attribute Based Encryption to prevent the data security. However, the security protection issues during the Attribute Based Encryption are considered as challenging task. In the traditional method of cloud data storage, the data are usually encrypted in the server side and then securely stored in remote server. Many researchers proposed many algorithms in which it makes easier for user's convenience so that it makes them fulfil their requirements. In, this proposed work detailed comparative study for data deduplication techniques in cloud storage are analysed. The results indicate that the cloud computing allows the users to perform the limited outsourcing performance of computational task with extraordinary server. Our proposed deduplication scheme enhances to improve a secured connection for attribute-based encryption for an emerging source to use and it proved the secured against the application system.
{"title":"Comparative Study and Secure Data Deduplication techniques for Cloud Computing storage","authors":"P. Malathi, S. Suganthidevi","doi":"10.1109/ICSES52305.2021.9633960","DOIUrl":"https://doi.org/10.1109/ICSES52305.2021.9633960","url":null,"abstract":"Cloud based communication is a tremendous growth in the world and which provides the services through virtualized resources with support of internet. In cloud computing, optimized storage techniques require to store the hugea mount of data. Performing the data deduplication technique over the encrypted data consider as the major challenging task. Eliminating the redundant data file from the optimized storage, which results the minimizing the band width and reduces the cost and disk usage. In Existing approach, the most of the research work are using Attribute Based Encryption to prevent the data security. However, the security protection issues during the Attribute Based Encryption are considered as challenging task. In the traditional method of cloud data storage, the data are usually encrypted in the server side and then securely stored in remote server. Many researchers proposed many algorithms in which it makes easier for user's convenience so that it makes them fulfil their requirements. In, this proposed work detailed comparative study for data deduplication techniques in cloud storage are analysed. The results indicate that the cloud computing allows the users to perform the limited outsourcing performance of computational task with extraordinary server. Our proposed deduplication scheme enhances to improve a secured connection for attribute-based encryption for an emerging source to use and it proved the secured against the application system.","PeriodicalId":6777,"journal":{"name":"2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES)","volume":"62 1","pages":"1-5"},"PeriodicalIF":0.0,"publicationDate":"2021-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84963048","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 : 2021-09-24DOI: 10.1109/ICSES52305.2021.9633789
Cheshta Kwatra, K. Gupta
Text Summarization is a widely researched and successful area of Natural Language Processing application. However, it remains limited to established languages such as English, French, etc. In this paper, we propose and compare extractive and abstractive summarization techniques for Hindi text documents. For either summarization, we first propose ward hierarchical agglomerative clustering. This is followed by the PageRank algorithm for extractive summarization while in abstractive summarization, we present an approach based on multi-sentence compression which only requires a POS tagger to generate Hindi text summaries.
{"title":"Extractive and Abstractive Summarization for Hindi Text using Hierarchical Clustering","authors":"Cheshta Kwatra, K. Gupta","doi":"10.1109/ICSES52305.2021.9633789","DOIUrl":"https://doi.org/10.1109/ICSES52305.2021.9633789","url":null,"abstract":"Text Summarization is a widely researched and successful area of Natural Language Processing application. However, it remains limited to established languages such as English, French, etc. In this paper, we propose and compare extractive and abstractive summarization techniques for Hindi text documents. For either summarization, we first propose ward hierarchical agglomerative clustering. This is followed by the PageRank algorithm for extractive summarization while in abstractive summarization, we present an approach based on multi-sentence compression which only requires a POS tagger to generate Hindi text summaries.","PeriodicalId":6777,"journal":{"name":"2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES)","volume":"23 1","pages":"1-6"},"PeriodicalIF":0.0,"publicationDate":"2021-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88295084","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 : 2021-09-24DOI: 10.1109/ICSES52305.2021.9633831
A. Keller, Anukul Pandey
The recognition/diagnosis of Parkinson's disease must be highly accurate to reduce the severity of the disorder with timely treatment. It is often seen that handwriting of the patient diminishes because it is tough to hold the pen/pencil due to muscle rigidity as the disease progresses. Men and women are neurologically different and so are the young and aged and thus respond differently to Parkinson's manifestation. Additionally, there is a significant link between the dominant hand of the person and the side of the body where the initial manifestation of the disease begins. This lays the foundation for research-based on gender, age and handedness (lateralization) to predict the disease. The HandPD dataset used here is inherently imbalanced. This gives rise to the issue of prediction model biasedness. The true nature of such a model is not quite revealed by the conventional accuracy alone. Thus, balanced accuracy is used to evaluate true efficiency. The technique proposed here alleviates model bias using hybrid resampling and extreme gradient boosting. It also explores the impact of features like age, gender and handedness on the mode efficiency. Experimental results of the technique proposed here yield the highest accuracy of 98.24%, a balanced accuracy of 98.14% with 100% sensitivity and 96.29% specificity when the age of the person is taken into account along with features extracted from the handwritten images.
{"title":"Hybrid Resampling and Xgboost Prediction Using Patient's Details as Features for Parkinson's Disease Detection","authors":"A. Keller, Anukul Pandey","doi":"10.1109/ICSES52305.2021.9633831","DOIUrl":"https://doi.org/10.1109/ICSES52305.2021.9633831","url":null,"abstract":"The recognition/diagnosis of Parkinson's disease must be highly accurate to reduce the severity of the disorder with timely treatment. It is often seen that handwriting of the patient diminishes because it is tough to hold the pen/pencil due to muscle rigidity as the disease progresses. Men and women are neurologically different and so are the young and aged and thus respond differently to Parkinson's manifestation. Additionally, there is a significant link between the dominant hand of the person and the side of the body where the initial manifestation of the disease begins. This lays the foundation for research-based on gender, age and handedness (lateralization) to predict the disease. The HandPD dataset used here is inherently imbalanced. This gives rise to the issue of prediction model biasedness. The true nature of such a model is not quite revealed by the conventional accuracy alone. Thus, balanced accuracy is used to evaluate true efficiency. The technique proposed here alleviates model bias using hybrid resampling and extreme gradient boosting. It also explores the impact of features like age, gender and handedness on the mode efficiency. Experimental results of the technique proposed here yield the highest accuracy of 98.24%, a balanced accuracy of 98.14% with 100% sensitivity and 96.29% specificity when the age of the person is taken into account along with features extracted from the handwritten images.","PeriodicalId":6777,"journal":{"name":"2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES)","volume":"4 1","pages":"1-7"},"PeriodicalIF":0.0,"publicationDate":"2021-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90105250","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 : 2021-09-24DOI: 10.1109/ICSES52305.2021.9633800
Y. L. Prasanna, Y. Tarakaram, Y. Mounika, R. Subramani
Recently, the use of large volumes of image data in many applications like internet has been increasing rapidly. So, to make an effective use of storage space and also bandwidth of the network, image compression is required. We have two kinds of image compression - one is lossy and other is lossless image compression. Lossy image compression produces a compressed image where quality of the image is maintained with some data loss. Lossy compression is widely used compared to lossless compression. Here, three lossy image compression techniques - Discrete Cosine Transform(DCT), Singular Value Decomposition (SVD) and Discrete Wavelet Transform(DWT) are used to perform image compression. These techniques are compared using some performance measures such as Peak Signal-to- Noise Ratio(PSNR), Compression Ratio(CR), Structural Similarity Index Measure(SSIM) and Mean Square Error(MSE).
{"title":"Comparison of Different Lossy Image Compression Techniques","authors":"Y. L. Prasanna, Y. Tarakaram, Y. Mounika, R. Subramani","doi":"10.1109/ICSES52305.2021.9633800","DOIUrl":"https://doi.org/10.1109/ICSES52305.2021.9633800","url":null,"abstract":"Recently, the use of large volumes of image data in many applications like internet has been increasing rapidly. So, to make an effective use of storage space and also bandwidth of the network, image compression is required. We have two kinds of image compression - one is lossy and other is lossless image compression. Lossy image compression produces a compressed image where quality of the image is maintained with some data loss. Lossy compression is widely used compared to lossless compression. Here, three lossy image compression techniques - Discrete Cosine Transform(DCT), Singular Value Decomposition (SVD) and Discrete Wavelet Transform(DWT) are used to perform image compression. These techniques are compared using some performance measures such as Peak Signal-to- Noise Ratio(PSNR), Compression Ratio(CR), Structural Similarity Index Measure(SSIM) and Mean Square Error(MSE).","PeriodicalId":6777,"journal":{"name":"2021 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES)","volume":"74 1","pages":"1-7"},"PeriodicalIF":0.0,"publicationDate":"2021-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86296354","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：