Rubeena khan - Academia.edu (original) (raw)

Papers by Rubeena khan

Research paper thumbnail of Improved skin cancer detection using CNN

International journal of health sciences

Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer.... more Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer. A variety of machine learning techniques have been developed in the past to detect such malignancies before they worsen. The goal of this article is to utilize a convolutional neural network to segment skin lesion images. The purpose of this study is to see how deep learning may be utilized to segment skin lesion photos. People may discover what skin diseases they may have, how to protect themselves from it, and what measures they can take early on to successfully treat the disease using Artificial Intelligence. Machine learning may be used to diagnose the problem and help us predict the result. The most widely used classification technology is the support vector machine. The discoveries might help doctors treat sickness early on and avoid further deterioration.

Research paper thumbnail of Skin cancer detection

International journal of health sciences

Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer.... more Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer. A variety of machine learning techniques have been developed in the past to detect such malignancies before they worsen. The goal of this article is to utilize a convolutional neural network to segment skin lesion images. The purpose of this study is to see how deep learning may be utilized to segment skin lesion photos. People may discover what skin diseases they may have, how to protect themselves from it, and what measures they can take early on to successfully treat the disease using Artificial Intelligence. Machine learning may be used to diagnose the problem and help us predict the result. The most widely used classification technology is the support vector machine. The discoveries might help doctors treat sickness early on and avoid further deterioration.

Research paper thumbnail of Improved skin cancer detection using CNN

International journal of health sciences

Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer.... more Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer. A variety of machine learning techniques have been developed in the past to detect such malignancies before they worsen. The goal of this article is to utilize a convolutional neural network to segment skin lesion images. The purpose of this study is to see how deep learning may be utilized to segment skin lesion photos. People may discover what skin diseases they may have, how to protect themselves from it, and what measures they can take early on to successfully treat the disease using Artificial Intelligence. Machine learning may be used to diagnose the problem and help us predict the result. The most widely used classification technology is the support vector machine. The discoveries might help doctors treat sickness early on and avoid further deterioration.

Research paper thumbnail of Web-Based Crowd Sentiment Analysis: Methods

Today, large population use social networking sites like Facebook, Twitter, LinkedIn etc. Through... more Today, large population use social networking sites like Facebook, Twitter, LinkedIn etc. Through social media, people share messages, photos. They also impart information about a particular event or specific situation. There is limited research on crowd management to handle a disaster. In this paper, we focus on Crowd Management using Sentiment Analysis as a tool for safety in some events or situations. People convey their emotion about crowd using social sites. Crowd-related issues encountered day to day life such as stations, shopping malls, and stadiums or some events like marriage which may cause congestion and due to that some people may be injured or causes death. Peoples post their sentiments through Twitter, LinkedIn etc. In this paper, tweets are collected from social networking site Twitter. These sentiments contain the list of expressions which have some polarity value. We use a rule-based algorithm for sentiment analysis. Public opinion can be classified into either pos...

Research paper thumbnail of 712 Rubeena Segmentation of Hindi Syllables Ending with Nasal Consonants Using ZCR and Dual Band Energy Ratio

In this paper a method for Hindi syllable segmentation for, syllables ending with nasal consonant... more In this paper a method for Hindi syllable segmentation for, syllables ending with nasal consonants is proposed. Syllable segmentation boundaries for words are computed using a combination of the zero-crossing-rate of the speech signal and a ratio of high to low frequency energy of the speech signal. Words comprising of syllables ending with nasal consonants are considered. The boundaries are computed by decomposing the signal into low and high frequency components using wavelet decomposition. A technique is proposed which uses the ratio of high to low frequency energy of the decomposed signal to compute the accurate syllable segmentation boundaries along with the ZCR function. The accuracy rate of syllable segmentation thus achieved is 88.67% for syllables ending with nasal consonants. INTRODUCTION Indian languages being syllable centered in nature produce natural sounding synthesized speech ,when syllables are used as the basic unit for synthesis [1]. Musfir etal. [2] suggest , tha...

Research paper thumbnail of Crowd Management using Sentiment Analysis

Today, many users are using Social networking sites such as Facebook, Twitter, LinkedIn etc. wher... more Today, many users are using Social networking sites such as Facebook, Twitter, LinkedIn etc. where user"s give their own opinion on particular event or specific situation. This paper focus on Crowd management and control using sentiment analysis. Congestion in crowded area is identified is noted through public opinions made on social networking sites. The public opinions are ambiguous and it is hard to analyse the situation manually or through simple algorithms. Peoples post their feeling through Twitter, LinkedIn etc. Their response for a specific situation is either positive, negative or neutral. The public opinions are then collected, processed and analysed using data mining techniques. In this paper, Sentiment analysis is done by rule-based algorithm. By analysis crowded area, we can move crowd to uncrowded area to avoid undesirable situation.

Research paper thumbnail of Raspberry Pi Home Automation with Voice Control

International Journal of Advance Research, Ideas and Innovations in Technology, 2017

Automation control means the use of various control systems (sensors) for operating equipment wit... more Automation control means the use of various control systems (sensors) for operating equipment without human interference. Home automation is exciting field when it is blow up with new technologies like voice control. It is computerization of the home or household activity. The suggested implementation of home automation bring under one control of lighting, heating, ventilation and it is fully controlled by using any smartphone through the particular android/iOS application and also with voice commands. Also, the Home appliances are controlled by voice command using Google speech API. The main advantage of this is small device can be part of internet so it is easy to communicate, manage and control without human interferences. Also, it provides the high degree of security, safety, comfort, and energy saving. The Raspberry Pi is small(85.60 mm x 56 mm x 21 mm), inexpensive(2,950 INR), portable, credit-size single board computer with support for a large number of peripherals like USB p...

Research paper thumbnail of Ranked Rule Based Approach for Sentiment Analysis

2018 International Conference on Recent Innovations in Electrical, Electronics & Communication Engineering (ICRIEECE), 2018

Today, large population use social networking sites like Facebook, Twitter, LinkedIn etc. Through... more Today, large population use social networking sites like Facebook, Twitter, LinkedIn etc. Through social media, people share messages, photos. They also impart information about a particular event or specific situation. There is limited research on crowd management to handle a disaster. We should focus on Crowd Management using Sentiment Analysis as a tool for safety in some events or situations. People convey their emotion about crowd using social sites. Crowd-related issues encountered day to day life such as stations, shopping malls, and stadiums or some events like marriage which may cause congestion and due to that some people may be injured or causes death. Peoples post their sentiments through Twitter, LinkedIn etc.In this paper, we consider traffic jam event where traffic will be able to move or will not be able to move. For this purpose, tweets are collected from social networking site Twitter. Human expressions are expressed through Natural Language Processing and then calculate polarity of sentiment using rule-based approach. User’s opinion is classified into positive, negative or neutral Sentiment. Polarity score of sentence is calculated through SND pattern. Users may enter false tweets which will decrease accuracy of system. To increase accuracy of system along with polarity score, we also consider polling based on user ranking in our proposed system.

Research paper thumbnail of Speech synthesis for hindi language using reduced database size and improved naturalness

Research paper thumbnail of Identification of Human Emotions from Speech Using CNN

Emotion is something which is derived from the environmental factors. In the writing of discourse... more Emotion is something which is derived from the environmental factors. In the writing of discourse of Speech Emotion Recognition (SER), numerous methods have been used to extricate emotions from speech. In the proposed framework for the acknowledgment of emotions from speech CNN algorithm is used. To identify emotions, feature extraction and preprocessing are done so that the undesirable clamor gets filtered out. These separated features were carried forward and sent to the CNN classifier model. The dataset will go through the extraction system which would settle on a choice with respect to the basic emotion perceives in the sound.

Research paper thumbnail of Secure Cloud Based Storage System with De-Duplication and Compression

Research paper thumbnail of Hindi Syllable Segmentation Using ZCR and Dual Band Energy Ratio

International Journal on Communications Antenna and Propagation (IRECAP), 2017

In this paper, a technique for Hindi syllable segmentation is proposed. Syllable segmentation bou... more In this paper, a technique for Hindi syllable segmentation is proposed. Syllable segmentation boundaries for words are first computed using the zero-crossing-rate of the speech signals. Words comprising of syllables ending with consonants and vowels are considered. The performance of the segmentation using a zero-crossing-rate algorithm can be further improved. The ZCR computed boundaries are optimized by decomposing the signal into low and high-frequency components using wavelet decomposition. A method is proposed which uses the ratio of the high to the low-frequency energy of the decomposed signal to compute the accurate syllable segmentation boundaries along with the ZCR function. The accuracy rate of syllable segmentation thus achieved is 96.02% for syllables ending with stop consonants and vowels.

Research paper thumbnail of Concatenative Speech Synthesis: A Review

International Journal of Computer Applications, 2016

The primary objective of this paper is to provide an overview of existing Concatenative Text-To-S... more The primary objective of this paper is to provide an overview of existing Concatenative Text-To-Speech synthesis techniques. Concatenative speech synthesis can be broadly categorized into three categories, Diphone Based, Corpus based and Hybrid. Diphone based speech synthesis relies on different signal processing techniques such as PSOLA, FD-PSOLA etc. These signal processing techniques introduce unwanted artifacts in the synthesized speech. The most popularly used method is the Unit selection synthesis which is a corpus based synthesis method. This method produces the most natural sounding synthetic speech.

Research paper thumbnail of Parallel Mining of Frequent Item sets using Map Reduce Technique: A Survey

In the task of data mining, the most important job is to find out frequent itemsets. Frequent ite... more In the task of data mining, the most important job is to find out frequent itemsets. Frequent itemsets are useful in various applications like Association rules and correlations. These systems are using some algorithms to find out frequent itemsets. But these parallel mining algorithms lack some features like automatic parallelization, well balancing the load, distribution of data on large number of clusters. So there is a need to study the parallel algorithms which will overcome the disadvantages of the existing system. In this paper a technique called fidoop is implemented, In this technique the mappers work independently as well as concurrently. This is done by decomposing the data across the mappers. Reducers work is to combine these jobs by developing small ultra metric trees. To show this fidoop technique on the various clusters is very delicate in distribution of data because different datasets are with different partition of data. This fidoop technique is also useful in hete...

Research paper thumbnail of Improved skin cancer detection using CNN

International journal of health sciences

Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer.... more Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer. A variety of machine learning techniques have been developed in the past to detect such malignancies before they worsen. The goal of this article is to utilize a convolutional neural network to segment skin lesion images. The purpose of this study is to see how deep learning may be utilized to segment skin lesion photos. People may discover what skin diseases they may have, how to protect themselves from it, and what measures they can take early on to successfully treat the disease using Artificial Intelligence. Machine learning may be used to diagnose the problem and help us predict the result. The most widely used classification technology is the support vector machine. The discoveries might help doctors treat sickness early on and avoid further deterioration.

Research paper thumbnail of Skin cancer detection

International journal of health sciences

Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer.... more Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer. A variety of machine learning techniques have been developed in the past to detect such malignancies before they worsen. The goal of this article is to utilize a convolutional neural network to segment skin lesion images. The purpose of this study is to see how deep learning may be utilized to segment skin lesion photos. People may discover what skin diseases they may have, how to protect themselves from it, and what measures they can take early on to successfully treat the disease using Artificial Intelligence. Machine learning may be used to diagnose the problem and help us predict the result. The most widely used classification technology is the support vector machine. The discoveries might help doctors treat sickness early on and avoid further deterioration.

Research paper thumbnail of Improved skin cancer detection using CNN

International journal of health sciences

Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer.... more Pollution, an unhealthy lifestyle, UV radiation, and other factors can contribute to skin cancer. A variety of machine learning techniques have been developed in the past to detect such malignancies before they worsen. The goal of this article is to utilize a convolutional neural network to segment skin lesion images. The purpose of this study is to see how deep learning may be utilized to segment skin lesion photos. People may discover what skin diseases they may have, how to protect themselves from it, and what measures they can take early on to successfully treat the disease using Artificial Intelligence. Machine learning may be used to diagnose the problem and help us predict the result. The most widely used classification technology is the support vector machine. The discoveries might help doctors treat sickness early on and avoid further deterioration.

Research paper thumbnail of Web-Based Crowd Sentiment Analysis: Methods

Today, large population use social networking sites like Facebook, Twitter, LinkedIn etc. Through... more Today, large population use social networking sites like Facebook, Twitter, LinkedIn etc. Through social media, people share messages, photos. They also impart information about a particular event or specific situation. There is limited research on crowd management to handle a disaster. In this paper, we focus on Crowd Management using Sentiment Analysis as a tool for safety in some events or situations. People convey their emotion about crowd using social sites. Crowd-related issues encountered day to day life such as stations, shopping malls, and stadiums or some events like marriage which may cause congestion and due to that some people may be injured or causes death. Peoples post their sentiments through Twitter, LinkedIn etc. In this paper, tweets are collected from social networking site Twitter. These sentiments contain the list of expressions which have some polarity value. We use a rule-based algorithm for sentiment analysis. Public opinion can be classified into either pos...

Research paper thumbnail of 712 Rubeena Segmentation of Hindi Syllables Ending with Nasal Consonants Using ZCR and Dual Band Energy Ratio

In this paper a method for Hindi syllable segmentation for, syllables ending with nasal consonant... more In this paper a method for Hindi syllable segmentation for, syllables ending with nasal consonants is proposed. Syllable segmentation boundaries for words are computed using a combination of the zero-crossing-rate of the speech signal and a ratio of high to low frequency energy of the speech signal. Words comprising of syllables ending with nasal consonants are considered. The boundaries are computed by decomposing the signal into low and high frequency components using wavelet decomposition. A technique is proposed which uses the ratio of high to low frequency energy of the decomposed signal to compute the accurate syllable segmentation boundaries along with the ZCR function. The accuracy rate of syllable segmentation thus achieved is 88.67% for syllables ending with nasal consonants. INTRODUCTION Indian languages being syllable centered in nature produce natural sounding synthesized speech ,when syllables are used as the basic unit for synthesis [1]. Musfir etal. [2] suggest , tha...

Research paper thumbnail of Crowd Management using Sentiment Analysis

Today, many users are using Social networking sites such as Facebook, Twitter, LinkedIn etc. wher... more Today, many users are using Social networking sites such as Facebook, Twitter, LinkedIn etc. where user"s give their own opinion on particular event or specific situation. This paper focus on Crowd management and control using sentiment analysis. Congestion in crowded area is identified is noted through public opinions made on social networking sites. The public opinions are ambiguous and it is hard to analyse the situation manually or through simple algorithms. Peoples post their feeling through Twitter, LinkedIn etc. Their response for a specific situation is either positive, negative or neutral. The public opinions are then collected, processed and analysed using data mining techniques. In this paper, Sentiment analysis is done by rule-based algorithm. By analysis crowded area, we can move crowd to uncrowded area to avoid undesirable situation.

Research paper thumbnail of Raspberry Pi Home Automation with Voice Control

International Journal of Advance Research, Ideas and Innovations in Technology, 2017

Automation control means the use of various control systems (sensors) for operating equipment wit... more Automation control means the use of various control systems (sensors) for operating equipment without human interference. Home automation is exciting field when it is blow up with new technologies like voice control. It is computerization of the home or household activity. The suggested implementation of home automation bring under one control of lighting, heating, ventilation and it is fully controlled by using any smartphone through the particular android/iOS application and also with voice commands. Also, the Home appliances are controlled by voice command using Google speech API. The main advantage of this is small device can be part of internet so it is easy to communicate, manage and control without human interferences. Also, it provides the high degree of security, safety, comfort, and energy saving. The Raspberry Pi is small(85.60 mm x 56 mm x 21 mm), inexpensive(2,950 INR), portable, credit-size single board computer with support for a large number of peripherals like USB p...

Research paper thumbnail of Ranked Rule Based Approach for Sentiment Analysis

2018 International Conference on Recent Innovations in Electrical, Electronics & Communication Engineering (ICRIEECE), 2018

Today, large population use social networking sites like Facebook, Twitter, LinkedIn etc. Through... more Today, large population use social networking sites like Facebook, Twitter, LinkedIn etc. Through social media, people share messages, photos. They also impart information about a particular event or specific situation. There is limited research on crowd management to handle a disaster. We should focus on Crowd Management using Sentiment Analysis as a tool for safety in some events or situations. People convey their emotion about crowd using social sites. Crowd-related issues encountered day to day life such as stations, shopping malls, and stadiums or some events like marriage which may cause congestion and due to that some people may be injured or causes death. Peoples post their sentiments through Twitter, LinkedIn etc.In this paper, we consider traffic jam event where traffic will be able to move or will not be able to move. For this purpose, tweets are collected from social networking site Twitter. Human expressions are expressed through Natural Language Processing and then calculate polarity of sentiment using rule-based approach. User’s opinion is classified into positive, negative or neutral Sentiment. Polarity score of sentence is calculated through SND pattern. Users may enter false tweets which will decrease accuracy of system. To increase accuracy of system along with polarity score, we also consider polling based on user ranking in our proposed system.

Research paper thumbnail of Speech synthesis for hindi language using reduced database size and improved naturalness

Research paper thumbnail of Identification of Human Emotions from Speech Using CNN

Emotion is something which is derived from the environmental factors. In the writing of discourse... more Emotion is something which is derived from the environmental factors. In the writing of discourse of Speech Emotion Recognition (SER), numerous methods have been used to extricate emotions from speech. In the proposed framework for the acknowledgment of emotions from speech CNN algorithm is used. To identify emotions, feature extraction and preprocessing are done so that the undesirable clamor gets filtered out. These separated features were carried forward and sent to the CNN classifier model. The dataset will go through the extraction system which would settle on a choice with respect to the basic emotion perceives in the sound.

Research paper thumbnail of Secure Cloud Based Storage System with De-Duplication and Compression

Research paper thumbnail of Hindi Syllable Segmentation Using ZCR and Dual Band Energy Ratio

International Journal on Communications Antenna and Propagation (IRECAP), 2017

In this paper, a technique for Hindi syllable segmentation is proposed. Syllable segmentation bou... more In this paper, a technique for Hindi syllable segmentation is proposed. Syllable segmentation boundaries for words are first computed using the zero-crossing-rate of the speech signals. Words comprising of syllables ending with consonants and vowels are considered. The performance of the segmentation using a zero-crossing-rate algorithm can be further improved. The ZCR computed boundaries are optimized by decomposing the signal into low and high-frequency components using wavelet decomposition. A method is proposed which uses the ratio of the high to the low-frequency energy of the decomposed signal to compute the accurate syllable segmentation boundaries along with the ZCR function. The accuracy rate of syllable segmentation thus achieved is 96.02% for syllables ending with stop consonants and vowels.

Research paper thumbnail of Concatenative Speech Synthesis: A Review

International Journal of Computer Applications, 2016

The primary objective of this paper is to provide an overview of existing Concatenative Text-To-S... more The primary objective of this paper is to provide an overview of existing Concatenative Text-To-Speech synthesis techniques. Concatenative speech synthesis can be broadly categorized into three categories, Diphone Based, Corpus based and Hybrid. Diphone based speech synthesis relies on different signal processing techniques such as PSOLA, FD-PSOLA etc. These signal processing techniques introduce unwanted artifacts in the synthesized speech. The most popularly used method is the Unit selection synthesis which is a corpus based synthesis method. This method produces the most natural sounding synthetic speech.

Research paper thumbnail of Parallel Mining of Frequent Item sets using Map Reduce Technique: A Survey

In the task of data mining, the most important job is to find out frequent itemsets. Frequent ite... more In the task of data mining, the most important job is to find out frequent itemsets. Frequent itemsets are useful in various applications like Association rules and correlations. These systems are using some algorithms to find out frequent itemsets. But these parallel mining algorithms lack some features like automatic parallelization, well balancing the load, distribution of data on large number of clusters. So there is a need to study the parallel algorithms which will overcome the disadvantages of the existing system. In this paper a technique called fidoop is implemented, In this technique the mappers work independently as well as concurrently. This is done by decomposing the data across the mappers. Reducers work is to combine these jobs by developing small ultra metric trees. To show this fidoop technique on the various clusters is very delicate in distribution of data because different datasets are with different partition of data. This fidoop technique is also useful in hete...