Since the day of an advent of the internet and the World Wide Web digital images started playing ... more Since the day of an advent of the internet and the World Wide Web digital images started playing an important role in exchanging ideas and sharing the information. Today it's hard to imagine the world of web without digital images right from simple personal web page to giant web applications. Mobile applications like WhatsApp also became the important means of the communication on which very often multimedia content is used to spread the word. But at the same time these digital images are vulnerable in the sense that, it can be easily manipulated to conceal or change the meaning. It could be dangerous, if these forgeries been done with malicious intentions. Hence it's very important to know and prove the authenticity of the digital images. From last few years, research community is contributing towards addressing this issue. In this survey paper we are presenting a review on copy-move forgery detection by dividing the image into the block of particular size and processing th...
Advances in Intelligent Systems and Computing, 2018
The online users' behavior is mainly categorized into two types: normal and suspicious behaviors.... more The online users' behavior is mainly categorized into two types: normal and suspicious behaviors. When any user is involved in group of activities that are deviated from normal behavior which is observed as misuse, damaging, intrusive, illegal money transfer, and online credit card fraud, then such users are become harmful. These users should be detected prior to the harmful activities. In this work, different online suspicious behavior patterns are studied and analyzed such as terrorists' activities, Web spam, e-mail spam, OSNs spam, fake reviews, malicious crawlers, banking frauds, hacking. Online credit card fraud is a suspicious activity and is taken into consideration for study. It is necessary to generate decision support system to predict the online credit card fraud as it will help to protect the illegal money transaction. The work represented in this research paper intends to find the best model to predict the online credit card fraud by using machine learning methods. Various machine learning methods such as logistic regression, SVM, kNN, Naïve Bayes, kernel SVM, decision tree, and random forest are implemented in order to predict the credit card fraud. The performance of the system is tested on the dataset which comes from Kaggle Web site. The performance of the algorithms is studied by considering different evaluation measures. Random forest algorithm turns out to be the best model in the prediction task. It has achieved accuracy of 99% and F-measure of 86%.
Machine learning inspires many practical solutions that maximize many resource utilization and pr... more Machine learning inspires many practical solutions that maximize many resource utilization and prolong the lifespan of a network. As wireless sensors network (WSNs) monitor dynamic environment that rapidly changes over time such behavior is either caused by the external factor or by initiated by the system designers. A comparative guide is provided to aid WSN designers to develop suitable machine learning solutions for appropriate application challenges. The security properties of sensors must be known before deploying the intelligent systems on critical infrastructure. This paper provides some steps for generating a comprehensive security model for sensors network as sensors network is not a traditional computing device hence existing security models are not applicable.
The traditional routing mechanism require the available path from the source to destination. Many... more The traditional routing mechanism require the available path from the source to destination. Many times, the available paths does not exist, also such path are multi-hop. The high mobility leads to frequent changes in the topology, thus causing the breakage of link. It causes the extra intermediate nodes to transfer the packets, which increase the transmission cost. The multi-hop transmission also increase the end to end delay. The two hop relay provide the better transmission over the network, reducing the end to end delay, packet delivery ratio, and transmission cost, etc. The two hop relay eliminates the condition of path connectivity of source to destination. The proposed algorithm provide the reliable transfer of the packet considering the various parameter.
Diagnostic imaging is a critical tool in healthcare sector. There are various modalities such as ... more Diagnostic imaging is a critical tool in healthcare sector. There are various modalities such as Magnetic resonance imaging (MRI), computed tomography (CT), digital mammography, and others, to provide an insight of subject’s body, noninvasively in order to facilitate stakeholders to take decision in diagnosis. Additionally, in medical research, these technologies has been playing centre role in most of the health care studies and experiments. Being a critical component in imaging systems, MRI system has been active area for researchers in computational intelligence and image processing. Due to advances in computing hardware and its easy availability, the performance of MRI system has been improved dramatically since its inception and is able to provide fast imaging, better resolution, immunity to artefacts and cheaper cost. One of the most important problems in image processing and analysis is segmentation and same is true for biomedical imaging. Various approaches has been reported...
Vehicular ad hoc networks (VANETs) enable vehicles to communicate with each other but require eff... more Vehicular ad hoc networks (VANETs) enable vehicles to communicate with each other but require efficient and robust routing protocols for their success. In this paper, exploit the infrastructure of roadside units (RSUs) to efficiently and reliably route packets in VANETs. Proposed system operates by using vehicles to carry and forward messages from a source vehicle to a nearby RSU and, if needed, route these messages through the RSU network and, finally send them from an RSU to the destination vehicle. Proposed system is mostly critical for users who are far apart and want to communicate using their vehicles onboard units.
To perform visualization on mobile devices different types of data may use like text, picture, ma... more To perform visualization on mobile devices different types of data may use like text, picture, maps, physical objects, abstract data etc. According to data visualization is categorized in two areas of visualization that is, scientific visualization and information visualization. Scientific visualization refers to some specific type of data like physical data and it is used for computer modeling and simulation. Information visualization refers to abstract data and used in business and finance, administration, digital media and other abstract concepts. The physical and abstract data is only one classification but there are others classification like static and dynamic data, structured and unstructured data, or hierarchical and non-hierarchical data classification. This paper is focus on information visualization of abstract data on mobile devices.
MANET consists of devices that run on batteries, having limited amount of energy and due to its s... more MANET consists of devices that run on batteries, having limited amount of energy and due to its self-configuring and dynamic nature, all operations are performed by the node itself. However, if any new technology and advancement are introduced in the MANET then the overhead of computation will also be performed by the individual nodes. All these computation will consume a lot of battery energy during the process of communication. In such scenario, DSR routing protocol is considered as base protocol and some modification are applied on it which acts into an efficient power aware DSR. In this paper, DSR is considered because it is one of the protocol which does not take into account energy, and once the dedicated path is established between source to destination then it will keep sending through that path until the link is broken. Whereas proposed DSR not only act as an energy efficient routing protocol but also shows an energy survival instinct. It establishes a path from source to d...
Text mining is a method that discovers interesting information in text documents. It is difficult... more Text mining is a method that discovers interesting information in text documents. It is difficult to find accurate feature in given text document for users to find what they want. Developing efficient feature extraction algorithms is highly needed to deal with high-dimensional data sets. Pattern mining techniques are used for finding appropriate features in both relevant and irrelevant documents. Pattern mining has been extensively studied in data mining communities for many years. This paper, discusses briefly five methods, i.e Fuzzy selfconstructing Feature Clustering method, Effective Pattern Discovery Technique, Learning Discriminative Phrase Pattern method, low-rank shared concept method, Relevant feature discovery model. This paper proposes a method improves the performance of the pattern mining, feature extraction method is used for information extraction from given text document. The proposed methodology for pattern mining from text documents using feature extraction method ...
Mobile computing environment are vulnerable to malicious attacks ranging from passive eavesdroppi... more Mobile computing environment are vulnerable to malicious attacks ranging from passive eavesdropping to an active interfering. In wireless network attacks can come from all the directions and target at any node. Compensation can include leaking secret information, message corruption and node masquerade. In mobile ad hoc networks security is hard due to the dynamic nature of the relationships between the participating nodes as well as the vulnerabilities and limitations of the wireless transmissions medium. In order to avoid such circumstance, there is need to develop new architecture and mechanisms to protect the wireless networks and mobile computing applications. All nodes in MANET must function as routers that discover and maintain routes to other nodes in the network. Due to MANET’s characteristics, it is a complicated job for IDS to detect routing attacks. Intrusion detection system plays an imperative role in network environment for security. Routing attacks may be launched by ...
International journal of simulation: systems, science & technology, 1970
With the massive improvements in the road transport network in India, the road safety is becoming... more With the massive improvements in the road transport network in India, the road safety is becoming the major concern of the authorities. The road safety is primarily defined by the motor skills of the driver. Nevertheless, the effects of road conditions on safety cannot be ignored. The authorities responsible for road maintenance are struggling to provide the higher maintenance of the roads to provide better, faster and safe transportation. Thus the demand for the research to automate the detection of the road conditions cannot be disregarded. A number of research attempts are carried out in order to detect the road condition based on the potholes. Nonetheless the detection process is highly time complex and makes the maintenance process delayed. Also, majority of the parallel research outcomes failed to measure multiple potholes in a single image and cannot distinguish the potholes based on the emergency of repair. Thus this work defines a newer dimension of pothole detection for road images, which contains higher number of potholes in a single image and makes the process faster by reducing the change of false detection. The false detection is usually the images in the dataset without having significant damages. Also, this work extracts the parameters for determining the potholes existence as the major outcome. Yet another outcome of this work is to classify the potholes based on the urgency of repair. The final outcome of the work is to automate the detection facility to provide a timely maintenance alert and deliver a better road condition in India.
2016 International Conference on Advances in Human Machine Interaction (HMI), 2016
Texting that is SMS is important function of anyMobile phone and we know that the mobile phone us... more Texting that is SMS is important function of anyMobile phone and we know that the mobile phone usage in theWorld is spreading rapidly and has gone through great changesdue to new developments and innovations in mobile phone technology. This paper is based on creating a messenger for the Differently-Abled set of Humans, who may not be in the positionof using mobile phones for messaging or any other kinds ofcommunicating devices, with the required comfort, we called it as application .In other words, messaging can be completely voicebased. The proposed Application is a Messaging System, which isVoice enabled.The application listens to your messages and then responds with voice commands by talking. The applicationconverts your text into voice and voice into text. For Android it isVoice-to-Text technology to listen to what you send and gets you connected with people.
International Journal of Innovative Research in Computer and Communication Engineering, Jan 8, 2015
Hashing function is the One of the most frequent way for finding the nearest match in the large d... more Hashing function is the One of the most frequent way for finding the nearest match in the large data sets. From the last decades number of researcher has been work on the hashing and focusing on the better approach than the existing one with respective their performance. In this paper presents a survey on different type of hash functions, different type of hashing method, hashing strategies and structural weakness of them or the limitation of them that in which kind of problem they are suitable and what they can't be used., also we are investigating the alternative approach for the mid-square hashing approach. The necessary data structures and algorithms are described, the expected performance is analyzed mathematically, and actual execution times are obtained and compared with alternative techniques. It shows that it provides the faster response time. Finally our method is intuitive and easy to implement.
This paper proposes a novel morphological approach for segmentation of document images in the Nat... more This paper proposes a novel morphological approach for segmentation of document images in the National Institute of Standards and Technology (NIST) Database. Document Image Analysis involves transformation of any information presented on paper into an equivalent symbolic representation accessible to computer information processing. For better storage and intelligent processing, information presented on paper is now being converted into electronic form. This needs processing of documents using image analysis algorithms. The morphological approach quantitatively describes operations effective for the shape of objects in an image. The mathematical morphology describes such operations by combinations of basic set operations between an image and a small object called a structuring element. The segmentation oriented geometrical features such as size, shape, contrast, or connectivity are efficiently dealt in mathematical morphology. In this paper, morphological approach is used for segmenting the document images. The performance of the proposed method has been evaluated using many text documents. The National Institute of Standards and Technology, Technology Administration, U. S. database of document images has been used in order to carry out experimental work. The image document database used is NIST DATABASE, Federal Register Document Image Database. NIST Special Database 25volume 1.(NISTIR 6245). The proposed technique has been tested for various documents like multi-column, multi-article, multi-font, multi-orientation and promising results are obtained. At the end the obtained performance is also compared with other well-known page segmentation algorithms.
It is a well known fact that, speech recognition systems perform well when the system is used in ... more It is a well known fact that, speech recognition systems perform well when the system is used in conditions similar to the one used to train the acoustic models. However, mismatches degrade the performance. In adverse environment, it is very difficult to predict the category of noise in advance in case of real world environmental noise and difficult to achieve environmental robustness. After doing rigorous experimental study it is observed that, a unique method is not available that will clean the noisy speech as well as preserve the quality which have been corrupted by real natural environmental (mixed) noise. It is also observed that only back-end techniques are not sufficient to improve the performance of a speech recognition system. It is necessary to implement performance improvement techniques at every step of back-end as well as front-end of the Automatic Speech Recognition (ASR) model. Current recognition systems solve this problem using a technique called adaptation. This study presents an experimental study that aims two points, first is to implement the hybrid method that will take care of clarifying the speech signal as much as possible with all combinations of filters and enhancement techniques. The second point is to develop a method for training all categories of noise that can adapt the acoustic models for a new environment that will help to improve the performance of the speech recognizer under real world environmental mismatched conditions. This experiment confirms that hybrid adaptation methods improve the ASR performance on both levels, (Signal-to-Noise Ratio) SNR improvement as well as word recognition accuracy in real world noisy environment.
Noise is ubiquitous in almost all acoustic environments. The speech signal, that is recorded by a... more Noise is ubiquitous in almost all acoustic environments. The speech signal, that is recorded by a microphone is generally infected by noise originating from various sources. Such contamination can change the characteristics of the speech signals and degrade the speech quality and intelligibility, thereby causing significant harm to human-to-machine communication systems. Noise detection and reduction for speech applications is often formulated as a digital filtering problem, where the clean speech estimation is obtained by passing the noisy speech through a linear filter. With such a formulation, the core issue of noise reduction becomes how to design an optimal filter that can significantly suppress noise without noticeable speech distortion. This paper focuses on voice activity detection, noise estimation, removal techniques and an optimal filter.
International Journal of Advancements in Computing Technology, 2010
The well known fact is that the performance of the Speech Recognition System degrades drastically... more The well known fact is that the performance of the Speech Recognition System degrades drastically in Adverse Environments. Additive noise is one of the major element of adverse noisy environment. Detecting voiced, un-voiced or silent speech signal in noisy environment is not an easy task. A voice activity detector (VAD) is effective when the noise is stationary; it often fails when the noise statistics change during speech presence. Moreover, accurate voice activity detection under very low signal-to-noise-ratio (SNR) conditions is not trivial. The noise estimate can have a major impact on the quality of the enhanced signal. If the noise estimate is too low, annoying residual noise will be audible, while if the noise estimate is too high, speech will be distorted resulting possibly in intelligibility loss. The different VAD methods noise tracking approaches need to implement to enhance the speech signals collected through microphone in human-computer interaction. Although such approaches might work satisfactorily in stationary noise (e.g., white noise), but they may not work well in more realistic environments (e.g., in a restaurant) where the spectral characteristics of the noise might be changing constantly. Hence there is a need to update the noise spectrum continuously over time and this can be done using noise-tracking algorithms. This paper presents some Voice Activity Detecting (VAD) and noise tracking approaches that will help to improve the performance of speech recognition system in adverse environment for humancomputer interaction.
International Journal of Computer Science and Information Technology, 2011
Considering the limitations of mobile devices like low bandwidth, less computation power, minimum... more Considering the limitations of mobile devices like low bandwidth, less computation power, minimum storage capacity etc it is not possible to store whole data for 3D visualization on mobile devices. Therefore to minimize the load of mobile devices there is use of server in case of remote 3D visualization on mobile devices (clients). For 3D visualization on mobile devices various techniques are used at server side as well as at mobile side for different purpose. Some techniques directly provides 3D visualization and some techniques are indirectly responsible for 3D visualization on mobile devices. Remote visualization on mobile devices includes various parameters and aspects in different techniques.
Since the day of an advent of the internet and the World Wide Web digital images started playing ... more Since the day of an advent of the internet and the World Wide Web digital images started playing an important role in exchanging ideas and sharing the information. Today it's hard to imagine the world of web without digital images right from simple personal web page to giant web applications. Mobile applications like WhatsApp also became the important means of the communication on which very often multimedia content is used to spread the word. But at the same time these digital images are vulnerable in the sense that, it can be easily manipulated to conceal or change the meaning. It could be dangerous, if these forgeries been done with malicious intentions. Hence it's very important to know and prove the authenticity of the digital images. From last few years, research community is contributing towards addressing this issue. In this survey paper we are presenting a review on copy-move forgery detection by dividing the image into the block of particular size and processing th...
Advances in Intelligent Systems and Computing, 2018
The online users' behavior is mainly categorized into two types: normal and suspicious behaviors.... more The online users' behavior is mainly categorized into two types: normal and suspicious behaviors. When any user is involved in group of activities that are deviated from normal behavior which is observed as misuse, damaging, intrusive, illegal money transfer, and online credit card fraud, then such users are become harmful. These users should be detected prior to the harmful activities. In this work, different online suspicious behavior patterns are studied and analyzed such as terrorists' activities, Web spam, e-mail spam, OSNs spam, fake reviews, malicious crawlers, banking frauds, hacking. Online credit card fraud is a suspicious activity and is taken into consideration for study. It is necessary to generate decision support system to predict the online credit card fraud as it will help to protect the illegal money transaction. The work represented in this research paper intends to find the best model to predict the online credit card fraud by using machine learning methods. Various machine learning methods such as logistic regression, SVM, kNN, Naïve Bayes, kernel SVM, decision tree, and random forest are implemented in order to predict the credit card fraud. The performance of the system is tested on the dataset which comes from Kaggle Web site. The performance of the algorithms is studied by considering different evaluation measures. Random forest algorithm turns out to be the best model in the prediction task. It has achieved accuracy of 99% and F-measure of 86%.
Machine learning inspires many practical solutions that maximize many resource utilization and pr... more Machine learning inspires many practical solutions that maximize many resource utilization and prolong the lifespan of a network. As wireless sensors network (WSNs) monitor dynamic environment that rapidly changes over time such behavior is either caused by the external factor or by initiated by the system designers. A comparative guide is provided to aid WSN designers to develop suitable machine learning solutions for appropriate application challenges. The security properties of sensors must be known before deploying the intelligent systems on critical infrastructure. This paper provides some steps for generating a comprehensive security model for sensors network as sensors network is not a traditional computing device hence existing security models are not applicable.
The traditional routing mechanism require the available path from the source to destination. Many... more The traditional routing mechanism require the available path from the source to destination. Many times, the available paths does not exist, also such path are multi-hop. The high mobility leads to frequent changes in the topology, thus causing the breakage of link. It causes the extra intermediate nodes to transfer the packets, which increase the transmission cost. The multi-hop transmission also increase the end to end delay. The two hop relay provide the better transmission over the network, reducing the end to end delay, packet delivery ratio, and transmission cost, etc. The two hop relay eliminates the condition of path connectivity of source to destination. The proposed algorithm provide the reliable transfer of the packet considering the various parameter.
Diagnostic imaging is a critical tool in healthcare sector. There are various modalities such as ... more Diagnostic imaging is a critical tool in healthcare sector. There are various modalities such as Magnetic resonance imaging (MRI), computed tomography (CT), digital mammography, and others, to provide an insight of subject’s body, noninvasively in order to facilitate stakeholders to take decision in diagnosis. Additionally, in medical research, these technologies has been playing centre role in most of the health care studies and experiments. Being a critical component in imaging systems, MRI system has been active area for researchers in computational intelligence and image processing. Due to advances in computing hardware and its easy availability, the performance of MRI system has been improved dramatically since its inception and is able to provide fast imaging, better resolution, immunity to artefacts and cheaper cost. One of the most important problems in image processing and analysis is segmentation and same is true for biomedical imaging. Various approaches has been reported...
Vehicular ad hoc networks (VANETs) enable vehicles to communicate with each other but require eff... more Vehicular ad hoc networks (VANETs) enable vehicles to communicate with each other but require efficient and robust routing protocols for their success. In this paper, exploit the infrastructure of roadside units (RSUs) to efficiently and reliably route packets in VANETs. Proposed system operates by using vehicles to carry and forward messages from a source vehicle to a nearby RSU and, if needed, route these messages through the RSU network and, finally send them from an RSU to the destination vehicle. Proposed system is mostly critical for users who are far apart and want to communicate using their vehicles onboard units.
To perform visualization on mobile devices different types of data may use like text, picture, ma... more To perform visualization on mobile devices different types of data may use like text, picture, maps, physical objects, abstract data etc. According to data visualization is categorized in two areas of visualization that is, scientific visualization and information visualization. Scientific visualization refers to some specific type of data like physical data and it is used for computer modeling and simulation. Information visualization refers to abstract data and used in business and finance, administration, digital media and other abstract concepts. The physical and abstract data is only one classification but there are others classification like static and dynamic data, structured and unstructured data, or hierarchical and non-hierarchical data classification. This paper is focus on information visualization of abstract data on mobile devices.
MANET consists of devices that run on batteries, having limited amount of energy and due to its s... more MANET consists of devices that run on batteries, having limited amount of energy and due to its self-configuring and dynamic nature, all operations are performed by the node itself. However, if any new technology and advancement are introduced in the MANET then the overhead of computation will also be performed by the individual nodes. All these computation will consume a lot of battery energy during the process of communication. In such scenario, DSR routing protocol is considered as base protocol and some modification are applied on it which acts into an efficient power aware DSR. In this paper, DSR is considered because it is one of the protocol which does not take into account energy, and once the dedicated path is established between source to destination then it will keep sending through that path until the link is broken. Whereas proposed DSR not only act as an energy efficient routing protocol but also shows an energy survival instinct. It establishes a path from source to d...
Text mining is a method that discovers interesting information in text documents. It is difficult... more Text mining is a method that discovers interesting information in text documents. It is difficult to find accurate feature in given text document for users to find what they want. Developing efficient feature extraction algorithms is highly needed to deal with high-dimensional data sets. Pattern mining techniques are used for finding appropriate features in both relevant and irrelevant documents. Pattern mining has been extensively studied in data mining communities for many years. This paper, discusses briefly five methods, i.e Fuzzy selfconstructing Feature Clustering method, Effective Pattern Discovery Technique, Learning Discriminative Phrase Pattern method, low-rank shared concept method, Relevant feature discovery model. This paper proposes a method improves the performance of the pattern mining, feature extraction method is used for information extraction from given text document. The proposed methodology for pattern mining from text documents using feature extraction method ...
Mobile computing environment are vulnerable to malicious attacks ranging from passive eavesdroppi... more Mobile computing environment are vulnerable to malicious attacks ranging from passive eavesdropping to an active interfering. In wireless network attacks can come from all the directions and target at any node. Compensation can include leaking secret information, message corruption and node masquerade. In mobile ad hoc networks security is hard due to the dynamic nature of the relationships between the participating nodes as well as the vulnerabilities and limitations of the wireless transmissions medium. In order to avoid such circumstance, there is need to develop new architecture and mechanisms to protect the wireless networks and mobile computing applications. All nodes in MANET must function as routers that discover and maintain routes to other nodes in the network. Due to MANET’s characteristics, it is a complicated job for IDS to detect routing attacks. Intrusion detection system plays an imperative role in network environment for security. Routing attacks may be launched by ...
International journal of simulation: systems, science & technology, 1970
With the massive improvements in the road transport network in India, the road safety is becoming... more With the massive improvements in the road transport network in India, the road safety is becoming the major concern of the authorities. The road safety is primarily defined by the motor skills of the driver. Nevertheless, the effects of road conditions on safety cannot be ignored. The authorities responsible for road maintenance are struggling to provide the higher maintenance of the roads to provide better, faster and safe transportation. Thus the demand for the research to automate the detection of the road conditions cannot be disregarded. A number of research attempts are carried out in order to detect the road condition based on the potholes. Nonetheless the detection process is highly time complex and makes the maintenance process delayed. Also, majority of the parallel research outcomes failed to measure multiple potholes in a single image and cannot distinguish the potholes based on the emergency of repair. Thus this work defines a newer dimension of pothole detection for road images, which contains higher number of potholes in a single image and makes the process faster by reducing the change of false detection. The false detection is usually the images in the dataset without having significant damages. Also, this work extracts the parameters for determining the potholes existence as the major outcome. Yet another outcome of this work is to classify the potholes based on the urgency of repair. The final outcome of the work is to automate the detection facility to provide a timely maintenance alert and deliver a better road condition in India.
2016 International Conference on Advances in Human Machine Interaction (HMI), 2016
Texting that is SMS is important function of anyMobile phone and we know that the mobile phone us... more Texting that is SMS is important function of anyMobile phone and we know that the mobile phone usage in theWorld is spreading rapidly and has gone through great changesdue to new developments and innovations in mobile phone technology. This paper is based on creating a messenger for the Differently-Abled set of Humans, who may not be in the positionof using mobile phones for messaging or any other kinds ofcommunicating devices, with the required comfort, we called it as application .In other words, messaging can be completely voicebased. The proposed Application is a Messaging System, which isVoice enabled.The application listens to your messages and then responds with voice commands by talking. The applicationconverts your text into voice and voice into text. For Android it isVoice-to-Text technology to listen to what you send and gets you connected with people.
International Journal of Innovative Research in Computer and Communication Engineering, Jan 8, 2015
Hashing function is the One of the most frequent way for finding the nearest match in the large d... more Hashing function is the One of the most frequent way for finding the nearest match in the large data sets. From the last decades number of researcher has been work on the hashing and focusing on the better approach than the existing one with respective their performance. In this paper presents a survey on different type of hash functions, different type of hashing method, hashing strategies and structural weakness of them or the limitation of them that in which kind of problem they are suitable and what they can't be used., also we are investigating the alternative approach for the mid-square hashing approach. The necessary data structures and algorithms are described, the expected performance is analyzed mathematically, and actual execution times are obtained and compared with alternative techniques. It shows that it provides the faster response time. Finally our method is intuitive and easy to implement.
This paper proposes a novel morphological approach for segmentation of document images in the Nat... more This paper proposes a novel morphological approach for segmentation of document images in the National Institute of Standards and Technology (NIST) Database. Document Image Analysis involves transformation of any information presented on paper into an equivalent symbolic representation accessible to computer information processing. For better storage and intelligent processing, information presented on paper is now being converted into electronic form. This needs processing of documents using image analysis algorithms. The morphological approach quantitatively describes operations effective for the shape of objects in an image. The mathematical morphology describes such operations by combinations of basic set operations between an image and a small object called a structuring element. The segmentation oriented geometrical features such as size, shape, contrast, or connectivity are efficiently dealt in mathematical morphology. In this paper, morphological approach is used for segmenting the document images. The performance of the proposed method has been evaluated using many text documents. The National Institute of Standards and Technology, Technology Administration, U. S. database of document images has been used in order to carry out experimental work. The image document database used is NIST DATABASE, Federal Register Document Image Database. NIST Special Database 25volume 1.(NISTIR 6245). The proposed technique has been tested for various documents like multi-column, multi-article, multi-font, multi-orientation and promising results are obtained. At the end the obtained performance is also compared with other well-known page segmentation algorithms.
It is a well known fact that, speech recognition systems perform well when the system is used in ... more It is a well known fact that, speech recognition systems perform well when the system is used in conditions similar to the one used to train the acoustic models. However, mismatches degrade the performance. In adverse environment, it is very difficult to predict the category of noise in advance in case of real world environmental noise and difficult to achieve environmental robustness. After doing rigorous experimental study it is observed that, a unique method is not available that will clean the noisy speech as well as preserve the quality which have been corrupted by real natural environmental (mixed) noise. It is also observed that only back-end techniques are not sufficient to improve the performance of a speech recognition system. It is necessary to implement performance improvement techniques at every step of back-end as well as front-end of the Automatic Speech Recognition (ASR) model. Current recognition systems solve this problem using a technique called adaptation. This study presents an experimental study that aims two points, first is to implement the hybrid method that will take care of clarifying the speech signal as much as possible with all combinations of filters and enhancement techniques. The second point is to develop a method for training all categories of noise that can adapt the acoustic models for a new environment that will help to improve the performance of the speech recognizer under real world environmental mismatched conditions. This experiment confirms that hybrid adaptation methods improve the ASR performance on both levels, (Signal-to-Noise Ratio) SNR improvement as well as word recognition accuracy in real world noisy environment.
Noise is ubiquitous in almost all acoustic environments. The speech signal, that is recorded by a... more Noise is ubiquitous in almost all acoustic environments. The speech signal, that is recorded by a microphone is generally infected by noise originating from various sources. Such contamination can change the characteristics of the speech signals and degrade the speech quality and intelligibility, thereby causing significant harm to human-to-machine communication systems. Noise detection and reduction for speech applications is often formulated as a digital filtering problem, where the clean speech estimation is obtained by passing the noisy speech through a linear filter. With such a formulation, the core issue of noise reduction becomes how to design an optimal filter that can significantly suppress noise without noticeable speech distortion. This paper focuses on voice activity detection, noise estimation, removal techniques and an optimal filter.
International Journal of Advancements in Computing Technology, 2010
The well known fact is that the performance of the Speech Recognition System degrades drastically... more The well known fact is that the performance of the Speech Recognition System degrades drastically in Adverse Environments. Additive noise is one of the major element of adverse noisy environment. Detecting voiced, un-voiced or silent speech signal in noisy environment is not an easy task. A voice activity detector (VAD) is effective when the noise is stationary; it often fails when the noise statistics change during speech presence. Moreover, accurate voice activity detection under very low signal-to-noise-ratio (SNR) conditions is not trivial. The noise estimate can have a major impact on the quality of the enhanced signal. If the noise estimate is too low, annoying residual noise will be audible, while if the noise estimate is too high, speech will be distorted resulting possibly in intelligibility loss. The different VAD methods noise tracking approaches need to implement to enhance the speech signals collected through microphone in human-computer interaction. Although such approaches might work satisfactorily in stationary noise (e.g., white noise), but they may not work well in more realistic environments (e.g., in a restaurant) where the spectral characteristics of the noise might be changing constantly. Hence there is a need to update the noise spectrum continuously over time and this can be done using noise-tracking algorithms. This paper presents some Voice Activity Detecting (VAD) and noise tracking approaches that will help to improve the performance of speech recognition system in adverse environment for humancomputer interaction.
International Journal of Computer Science and Information Technology, 2011
Considering the limitations of mobile devices like low bandwidth, less computation power, minimum... more Considering the limitations of mobile devices like low bandwidth, less computation power, minimum storage capacity etc it is not possible to store whole data for 3D visualization on mobile devices. Therefore to minimize the load of mobile devices there is use of server in case of remote 3D visualization on mobile devices (clients). For 3D visualization on mobile devices various techniques are used at server side as well as at mobile side for different purpose. Some techniques directly provides 3D visualization and some techniques are indirectly responsible for 3D visualization on mobile devices. Remote visualization on mobile devices includes various parameters and aspects in different techniques.
Uploads
Papers by Vilas Thakare