Pub Date : 2021-08-26DOI: 10.1109/SPIN52536.2021.9566067
A. Agarwal, R. Tiwari, Vikas Khullar, R. Kaushal
Machine learning techniques enable systems to learn Important representations from input Image data. Convolutional neural networks (CNNs) are a specific implementation of machine learning techniques and are able to create expressive representations from the input image. Hence CNNs are well suited for image processing operations such as classification, clustering, and object detection, etc. The creation of a new effectual deep CNN model involves an extensive training phase. This requires very large datasets, huge computation environments, and longer execution time. Several established deep CNNs are readily available. These networks are pre-trained on massive databases of images. VGG, ResNet, and InceptionResNetVZ are the leading pre-trained CNN models currently being used in numerous image-processing studies. Possibly we can transfer knowledge learned from such models in order to address challenges in different domains. This can be achieved by repurposing a deep CNN model as a feature generator to produce effective features for content based information retrieval applications. This research work proposes a technique for recognizing fish using deep convolutional neural networks such as ResNet-50, InceptionResNetVZ, and VGG16 that have been pre-trained using transfer learning.
{"title":"Transfer Learning Inspired Fish Species Classification","authors":"A. Agarwal, R. Tiwari, Vikas Khullar, R. Kaushal","doi":"10.1109/SPIN52536.2021.9566067","DOIUrl":"https://doi.org/10.1109/SPIN52536.2021.9566067","url":null,"abstract":"Machine learning techniques enable systems to learn Important representations from input Image data. Convolutional neural networks (CNNs) are a specific implementation of machine learning techniques and are able to create expressive representations from the input image. Hence CNNs are well suited for image processing operations such as classification, clustering, and object detection, etc. The creation of a new effectual deep CNN model involves an extensive training phase. This requires very large datasets, huge computation environments, and longer execution time. Several established deep CNNs are readily available. These networks are pre-trained on massive databases of images. VGG, ResNet, and InceptionResNetVZ are the leading pre-trained CNN models currently being used in numerous image-processing studies. Possibly we can transfer knowledge learned from such models in order to address challenges in different domains. This can be achieved by repurposing a deep CNN model as a feature generator to produce effective features for content based information retrieval applications. This research work proposes a technique for recognizing fish using deep convolutional neural networks such as ResNet-50, InceptionResNetVZ, and VGG16 that have been pre-trained using transfer learning.","PeriodicalId":343177,"journal":{"name":"2021 8th International Conference on Signal Processing and Integrated Networks (SPIN)","volume":"263 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127704476","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 : 2021-08-26DOI: 10.1109/SPIN52536.2021.9566015
Akanksha Madduri, Sai Sushma Adusumalli, Honey Sri Katragadda, Mohith Krishna Reddy Dontireddy, Pallikonda Sarah Suhasini
Breast Cancer is one of the mostly encountered cancers among women which involve the age group of 60-80 years mostly. The traditional methodology involves use of mammogram scan followed by various other clinical tests for assuring cancer prevailing in the body manually, which involves mistakes and delay in detection. Many times, it is detected using the biopsy method where tissue removed from the breast is studied under a microscope. This entire process is done by the histopathologies, and if he is not well trained, it may lead to wrong diagnosis. In order to improve the diagnosis by proper detection, automatic analysis of histopathology images has helped the pathologists in efficient diagnosis. Recently the Convolutional neural networks (CNN) have become a preferred deep learning method for breast cancer classification. In this paper, we have proposed CNN architecture based on Local Binary Pattern (LBP) images as input and then compare their classification results by a standard CNN based on origin images as input. Here, classification approach is proposed for automatic classification into either moderate stage or mild stage of cancer. An image dataset of 100 images is used in this approach and 80% of the dataset is used for training and the rest 20% of the images used for testing. 100% classification accuracy is obtained with CNN architecture. The results are compared with various state-of-art machine learning models.
{"title":"Classification of Breast Cancer Histopathological Images using Convolutional Neural Networks","authors":"Akanksha Madduri, Sai Sushma Adusumalli, Honey Sri Katragadda, Mohith Krishna Reddy Dontireddy, Pallikonda Sarah Suhasini","doi":"10.1109/SPIN52536.2021.9566015","DOIUrl":"https://doi.org/10.1109/SPIN52536.2021.9566015","url":null,"abstract":"Breast Cancer is one of the mostly encountered cancers among women which involve the age group of 60-80 years mostly. The traditional methodology involves use of mammogram scan followed by various other clinical tests for assuring cancer prevailing in the body manually, which involves mistakes and delay in detection. Many times, it is detected using the biopsy method where tissue removed from the breast is studied under a microscope. This entire process is done by the histopathologies, and if he is not well trained, it may lead to wrong diagnosis. In order to improve the diagnosis by proper detection, automatic analysis of histopathology images has helped the pathologists in efficient diagnosis. Recently the Convolutional neural networks (CNN) have become a preferred deep learning method for breast cancer classification. In this paper, we have proposed CNN architecture based on Local Binary Pattern (LBP) images as input and then compare their classification results by a standard CNN based on origin images as input. Here, classification approach is proposed for automatic classification into either moderate stage or mild stage of cancer. An image dataset of 100 images is used in this approach and 80% of the dataset is used for training and the rest 20% of the images used for testing. 100% classification accuracy is obtained with CNN architecture. The results are compared with various state-of-art machine learning models.","PeriodicalId":343177,"journal":{"name":"2021 8th International Conference on Signal Processing and Integrated Networks (SPIN)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127998000","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 : 2021-08-26DOI: 10.1109/SPIN52536.2021.9565960
N. Sharma, Anubhav Kumar, A. De, R. K. Jain
A compact and dual-band circularly polarized antenna with resonant frequency of 5.8GHz(ISM) and 7.6 GHz is proposed for biomedical, satellite and specific UWB applications. The 10 dB impedance bandwidth (IBW) of the antenna varies from 5.55 GHz to 5.94 GHz and 6.78 GHz to 8.78 GHz. The tilted arc-shaped radiator is used to perturb the current which is responsible for lower frequency band as well as circular polarization with the axial ratio extending from 5.77 GHz to 5.93 GHz, which covers 41% of the lower frequency band. The antenna is analyzed for wearable applications on a three-layer skin phantom model and the SAR value obtained is 0.2 W/Kg with Source power of 10mW, which is below the maximum permissible limit of 1.6W/Kg.
{"title":"Circularly Polarized Antenna for ISM (5.8 GHz), Satellite Communications and UWB Applications","authors":"N. Sharma, Anubhav Kumar, A. De, R. K. Jain","doi":"10.1109/SPIN52536.2021.9565960","DOIUrl":"https://doi.org/10.1109/SPIN52536.2021.9565960","url":null,"abstract":"A compact and dual-band circularly polarized antenna with resonant frequency of 5.8GHz(ISM) and 7.6 GHz is proposed for biomedical, satellite and specific UWB applications. The 10 dB impedance bandwidth (IBW) of the antenna varies from 5.55 GHz to 5.94 GHz and 6.78 GHz to 8.78 GHz. The tilted arc-shaped radiator is used to perturb the current which is responsible for lower frequency band as well as circular polarization with the axial ratio extending from 5.77 GHz to 5.93 GHz, which covers 41% of the lower frequency band. The antenna is analyzed for wearable applications on a three-layer skin phantom model and the SAR value obtained is 0.2 W/Kg with Source power of 10mW, which is below the maximum permissible limit of 1.6W/Kg.","PeriodicalId":343177,"journal":{"name":"2021 8th International Conference on Signal Processing and Integrated Networks (SPIN)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121457799","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 : 2021-08-26DOI: 10.1109/SPIN52536.2021.9566089
Vikas Rattan, Poonam Panwar, R. Mittal, Jaiteg Singh, Varun Malik
Approximating the budget of software is continuously vital problem for the software analysts to agree on the development of a project. Diverse methods for estimating the budget of software are used in literature based on the preferred reliability level as per client’s demand. Based on the literature a close correlation is found between software budget and its reliability. Hence, the architectural design of software is studied here using Discrete Time Markovian Chain (DTMC) for approximating its reliability and its association with software budget. Primarily, a conventional cost model known as generalized software cost model is used for evaluation of the software budget for eleven datasets obtained from literature. The proposed methodology is afterward used for approximation of the software budget and its overall reliability. The proposed methodology is then used to solve single objective problems of minimizing cost of software keeping reliability as constraint and maximizing reliability by keeping cost as constraint. At final, the proposed approach is used to analyse the software cost and reliability trade-off to provide best multi-objective solution to software clients.
{"title":"Forecasting the Budget Required for Software under Development","authors":"Vikas Rattan, Poonam Panwar, R. Mittal, Jaiteg Singh, Varun Malik","doi":"10.1109/SPIN52536.2021.9566089","DOIUrl":"https://doi.org/10.1109/SPIN52536.2021.9566089","url":null,"abstract":"Approximating the budget of software is continuously vital problem for the software analysts to agree on the development of a project. Diverse methods for estimating the budget of software are used in literature based on the preferred reliability level as per client’s demand. Based on the literature a close correlation is found between software budget and its reliability. Hence, the architectural design of software is studied here using Discrete Time Markovian Chain (DTMC) for approximating its reliability and its association with software budget. Primarily, a conventional cost model known as generalized software cost model is used for evaluation of the software budget for eleven datasets obtained from literature. The proposed methodology is afterward used for approximation of the software budget and its overall reliability. The proposed methodology is then used to solve single objective problems of minimizing cost of software keeping reliability as constraint and maximizing reliability by keeping cost as constraint. At final, the proposed approach is used to analyse the software cost and reliability trade-off to provide best multi-objective solution to software clients.","PeriodicalId":343177,"journal":{"name":"2021 8th International Conference on Signal Processing and Integrated Networks (SPIN)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128434871","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 : 2021-08-26DOI: 10.1109/SPIN52536.2021.9565950
Soumyo Das, A. Jose, Sai Ashish Kanna, Moby S. Philip, Anirudh Kumar
In this paper, the vehicle motion control for an automated angular reverse parking has been formulated and designed to track the planned path during auto angular park assist mode and aided in parking the vehicle diagonally in the parking slot without any side or rear collision. The diagonal reverse parking is one of the complicated functionalities and the quintessential features of the angular park assist system described in this paper helps the drivers for a successful maneuver during reverse parking. The architecture of the designed angular parking system which includes components such as situation assessment, mode manager, path planning, and controllers are formulated in enterprise architect. The trajectory for angular parking is designed for single maneuvering based on geometry and dynamics of vehicle maneuver which includes combination of circular, straight path during forward and reverse maneuver mode. The composite reverse motion control, including longitudinal support with braking to lateral controller, has been designed to follow the designed diagonal path precisely. The look rear concept based lateral motion control has been proposed in this paper to perform a successful parking while tracking planned trajectory. The objective of this research is to aid system to navigate host vehicle in a parking zone and help the system to follow planned trajectory precisely by minimizing the point-to-point positional error. The performance of proposed angular parking-controlled system is validated with kinematic vehicle model of Carmaker in loop while evaluating against pre-defined key performance indices of an angular parking.
{"title":"Design of Collision Free Automated Angular Parking","authors":"Soumyo Das, A. Jose, Sai Ashish Kanna, Moby S. Philip, Anirudh Kumar","doi":"10.1109/SPIN52536.2021.9565950","DOIUrl":"https://doi.org/10.1109/SPIN52536.2021.9565950","url":null,"abstract":"In this paper, the vehicle motion control for an automated angular reverse parking has been formulated and designed to track the planned path during auto angular park assist mode and aided in parking the vehicle diagonally in the parking slot without any side or rear collision. The diagonal reverse parking is one of the complicated functionalities and the quintessential features of the angular park assist system described in this paper helps the drivers for a successful maneuver during reverse parking. The architecture of the designed angular parking system which includes components such as situation assessment, mode manager, path planning, and controllers are formulated in enterprise architect. The trajectory for angular parking is designed for single maneuvering based on geometry and dynamics of vehicle maneuver which includes combination of circular, straight path during forward and reverse maneuver mode. The composite reverse motion control, including longitudinal support with braking to lateral controller, has been designed to follow the designed diagonal path precisely. The look rear concept based lateral motion control has been proposed in this paper to perform a successful parking while tracking planned trajectory. The objective of this research is to aid system to navigate host vehicle in a parking zone and help the system to follow planned trajectory precisely by minimizing the point-to-point positional error. The performance of proposed angular parking-controlled system is validated with kinematic vehicle model of Carmaker in loop while evaluating against pre-defined key performance indices of an angular parking.","PeriodicalId":343177,"journal":{"name":"2021 8th International Conference on Signal Processing and Integrated Networks (SPIN)","volume":"138 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128469464","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 : 2021-08-26DOI: 10.1109/SPIN52536.2021.9565986
Anju Rani, V. Arora, K. Sekhar, R. Mulaveesala
Frequency modulated thermography (FMT) is an efficient thermographic technique for quantitative analysis of defects in any material. The paper presents analytical solution of heat transfer in a finite thickness sample with flat bottom hole defects located at different lateral dimensions. The carbon fibre reinforced polymer (CFRP) sample is subjected to frequency modulated thermal excitation and temperature variations are evaluated for defect detection analysis. The computed analytical solutions for different defect depths have been shown to agree with corresponding simulation results for CFRP sample. The present work highlights defect detection capability of FMT technique using matched filter approach.
{"title":"Analytical Study of Frequency Modulated Thermography for Defect Estimation in Carbon Fibre Reinforced Polymer","authors":"Anju Rani, V. Arora, K. Sekhar, R. Mulaveesala","doi":"10.1109/SPIN52536.2021.9565986","DOIUrl":"https://doi.org/10.1109/SPIN52536.2021.9565986","url":null,"abstract":"Frequency modulated thermography (FMT) is an efficient thermographic technique for quantitative analysis of defects in any material. The paper presents analytical solution of heat transfer in a finite thickness sample with flat bottom hole defects located at different lateral dimensions. The carbon fibre reinforced polymer (CFRP) sample is subjected to frequency modulated thermal excitation and temperature variations are evaluated for defect detection analysis. The computed analytical solutions for different defect depths have been shown to agree with corresponding simulation results for CFRP sample. The present work highlights defect detection capability of FMT technique using matched filter approach.","PeriodicalId":343177,"journal":{"name":"2021 8th International Conference on Signal Processing and Integrated Networks (SPIN)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114523984","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 : 2021-08-26DOI: 10.1109/SPIN52536.2021.9566147
G. Prasad, Akriti Dikshit, S. Lalitha
With the onset of Covid-19, interactions between humans and machines have increased at a rapid rate. Helping the machine identify the emotion and sentiment of the user plays a key role in making these interactions feel more natural. To do so, existing models for Speech Emotion Recognition (SER) and Sentiment Analysis (SA) focus on the detection of either only emotion or sentiment on acted databases. Unlike these existing works, this work presents a simple model with a comparatively small speech feature vector, to detect both emotion and sentiment from the spontaneous database, Multimodal Emotion Lines Dataset (MELD). This contains voice samples similar to those in a real-time environment. Speech features such as Mel Frequency Cepstral Coefficients (MFCC), Entropy, Teager Energy Operator have been extracted from the voice samples and are classified using Logit Boost, Logistic and Multiclass classifier. The performance of the model is improved by using feature selection techniques such as Backward elimination and Gaussian distribution coefficients. The proposed model is simple, and the results are comparable to existing work on the MELD database.
{"title":"Sentiment and Emotion Analysis for Effective Human-Machine Interaction during Covid-19 Pandemic","authors":"G. Prasad, Akriti Dikshit, S. Lalitha","doi":"10.1109/SPIN52536.2021.9566147","DOIUrl":"https://doi.org/10.1109/SPIN52536.2021.9566147","url":null,"abstract":"With the onset of Covid-19, interactions between humans and machines have increased at a rapid rate. Helping the machine identify the emotion and sentiment of the user plays a key role in making these interactions feel more natural. To do so, existing models for Speech Emotion Recognition (SER) and Sentiment Analysis (SA) focus on the detection of either only emotion or sentiment on acted databases. Unlike these existing works, this work presents a simple model with a comparatively small speech feature vector, to detect both emotion and sentiment from the spontaneous database, Multimodal Emotion Lines Dataset (MELD). This contains voice samples similar to those in a real-time environment. Speech features such as Mel Frequency Cepstral Coefficients (MFCC), Entropy, Teager Energy Operator have been extracted from the voice samples and are classified using Logit Boost, Logistic and Multiclass classifier. The performance of the model is improved by using feature selection techniques such as Backward elimination and Gaussian distribution coefficients. The proposed model is simple, and the results are comparable to existing work on the MELD database.","PeriodicalId":343177,"journal":{"name":"2021 8th International Conference on Signal Processing and Integrated Networks (SPIN)","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127340067","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 : 2021-08-26DOI: 10.1109/SPIN52536.2021.9566079
V. Kukreja, Anupam Baliyan, Vikas Salonki, R. Kaushal
Detection of plant crop diseases has become an active field of research day by day due to increasing the demand for such systems and techniques as crop diseases are now become a common part of agriculture. Focusing on this demand and need, we have developed a Convolutional neural network (CNN)-based Deep learning (DL) multi-classification model which classifies the total of 900 real-time collected images of potato crop plants having healthy and potato blight (PB) disease images based on their PB disease severity level, along with this binary classification has also been done to simply classify the healthy and disease crop leaf. A total of four disease severity levels have been taken into account which resulted in a binary classification accuracy of 90.77% and 94.77% of best multi-classification accuracy. This work will be a great contribution in the field of potato disease recognition and detection using DL approaches.
{"title":"Potato Blight: Deep Learning Model for Binary and Multi-Classification","authors":"V. Kukreja, Anupam Baliyan, Vikas Salonki, R. Kaushal","doi":"10.1109/SPIN52536.2021.9566079","DOIUrl":"https://doi.org/10.1109/SPIN52536.2021.9566079","url":null,"abstract":"Detection of plant crop diseases has become an active field of research day by day due to increasing the demand for such systems and techniques as crop diseases are now become a common part of agriculture. Focusing on this demand and need, we have developed a Convolutional neural network (CNN)-based Deep learning (DL) multi-classification model which classifies the total of 900 real-time collected images of potato crop plants having healthy and potato blight (PB) disease images based on their PB disease severity level, along with this binary classification has also been done to simply classify the healthy and disease crop leaf. A total of four disease severity levels have been taken into account which resulted in a binary classification accuracy of 90.77% and 94.77% of best multi-classification accuracy. This work will be a great contribution in the field of potato disease recognition and detection using DL approaches.","PeriodicalId":343177,"journal":{"name":"2021 8th International Conference on Signal Processing and Integrated Networks (SPIN)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127072595","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 : 2021-08-26DOI: 10.1109/SPIN52536.2021.9566084
Saumendra Kumar Mohapatra, Rashmita Khilar, Abhishek Das, M. Mohanty
Cardiac data classification is an emerging research area in recent days. Machine learning-based automatic classification model is one of the essential aspects for the diagnosis of cardiac disease. The performance of a model can be improved by combining multiple models to solve a single problem. In this work, the authors have adopted a modified gradient boosting ensemble learning-based classifier for classifying the cardiac data collected from the UCI machine learning repository. The data set contains the samples of 303 patients with 13 attributes related to cardiac symptoms. The classification is done by using two types of gradient boosting ensemble classifier. In the first step, the classification is performed with a fixed learning rate of 0.01 for every tree. Further to improve the performance the learning rate is changed for each tree. From the result, it is observed that the accuracy is increasing with variation in learning rate. 91% accuracy is observed while the learning rate of 0.81 is considered. The performance is compared with the earlier works and is observed that the proposed model is providing a better result.
{"title":"Design of Gradient Boosting Ensemble Classifier with Variation of Learning Rate for Automated Cardiac Data Classification","authors":"Saumendra Kumar Mohapatra, Rashmita Khilar, Abhishek Das, M. Mohanty","doi":"10.1109/SPIN52536.2021.9566084","DOIUrl":"https://doi.org/10.1109/SPIN52536.2021.9566084","url":null,"abstract":"Cardiac data classification is an emerging research area in recent days. Machine learning-based automatic classification model is one of the essential aspects for the diagnosis of cardiac disease. The performance of a model can be improved by combining multiple models to solve a single problem. In this work, the authors have adopted a modified gradient boosting ensemble learning-based classifier for classifying the cardiac data collected from the UCI machine learning repository. The data set contains the samples of 303 patients with 13 attributes related to cardiac symptoms. The classification is done by using two types of gradient boosting ensemble classifier. In the first step, the classification is performed with a fixed learning rate of 0.01 for every tree. Further to improve the performance the learning rate is changed for each tree. From the result, it is observed that the accuracy is increasing with variation in learning rate. 91% accuracy is observed while the learning rate of 0.81 is considered. The performance is compared with the earlier works and is observed that the proposed model is providing a better result.","PeriodicalId":343177,"journal":{"name":"2021 8th International Conference on Signal Processing and Integrated Networks (SPIN)","volume":"2023 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122102566","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 : 2021-08-26DOI: 10.1109/SPIN52536.2021.9566005
Aniket Kumar, M. Madaan, Shubham Kumar, Aniket Saha, Suman Yadav
Communication is a basic requirement of an individual to exchange feelings, thoughts, and ideas, but the hearing and speech impaired community finds it difficult to interact with the vast majority of people. Sign language facilitates communication between the hearing and speech impaired person and the rest of society. The Rights of Persons with Disabilities (RPWD) Act, 2016, was also passed by the Indian government, which acknowledges Indian Sign Language (ISL) and mandates the use of sign language interpreters in all government-aided organizations and the public sector proceedings. Unfortunately, a large percentage of the Indian population is not familiar with the semantics of the gestures associated with ISL. To bridge this communication gap, this paper proposes a model to identify and classify Indian Sign Language gestures in real-time using Convolutional Neural Networks (CNN). The model has been developed using OpenCV and Keras implementation of CNNs and aims to classify 36 ISL gestures representing 0-9 numbers and A-Z alphabets by converting them to their text equivalents. The dataset created and used consists of 300 images for each gesture which were fed into the CNN model for training and testing purposes. The proposed model was successfully implemented and achieved 99.91% accuracy for the test images.
{"title":"Indian Sign Language Gesture Recognition in Real-Time using Convolutional Neural Networks","authors":"Aniket Kumar, M. Madaan, Shubham Kumar, Aniket Saha, Suman Yadav","doi":"10.1109/SPIN52536.2021.9566005","DOIUrl":"https://doi.org/10.1109/SPIN52536.2021.9566005","url":null,"abstract":"Communication is a basic requirement of an individual to exchange feelings, thoughts, and ideas, but the hearing and speech impaired community finds it difficult to interact with the vast majority of people. Sign language facilitates communication between the hearing and speech impaired person and the rest of society. The Rights of Persons with Disabilities (RPWD) Act, 2016, was also passed by the Indian government, which acknowledges Indian Sign Language (ISL) and mandates the use of sign language interpreters in all government-aided organizations and the public sector proceedings. Unfortunately, a large percentage of the Indian population is not familiar with the semantics of the gestures associated with ISL. To bridge this communication gap, this paper proposes a model to identify and classify Indian Sign Language gestures in real-time using Convolutional Neural Networks (CNN). The model has been developed using OpenCV and Keras implementation of CNNs and aims to classify 36 ISL gestures representing 0-9 numbers and A-Z alphabets by converting them to their text equivalents. The dataset created and used consists of 300 images for each gesture which were fed into the CNN model for training and testing purposes. The proposed model was successfully implemented and achieved 99.91% accuracy for the test images.","PeriodicalId":343177,"journal":{"name":"2021 8th International Conference on Signal Processing and Integrated Networks (SPIN)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123544573","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}