Pune, India
Pune, India

Time filter

Source Type

Dhanashri D.,AISSMS IOIT | Dhonde S.B.,AISSMS IOIT
Advances in Intelligent Systems and Computing | Year: 2017

Speech recognition is the process of converting speech signals into words. For acoustic modeling HMM-GMM is used for many years. For GMM, it requires assumptions near the data distribution for calculating probabilities. For removing this limitation, GMM is replaced by DNN in acoustic model. Deep neural networks are the feed forward neural networks having more than one or multiple layers of hidden units. In this work, we have presented the isolated word speech recognition system using acoustic model of HMM and DNN. We are using Deep Belief Network pre-training algorithm for initializing deep neural networks. DBN is a multilayer generative probabilistic model with large number of stochastic binary units. The features used are the mel-frequency cepstrum coefficients (MFCC). Experimental results are calculated on TI digits database. Proposed system has achieved 86.06% accuracy on TI digits database. System accuracy can be further increased by increasing the number of hidden units. © Springer Science+Business Media Singapore 2017.


Shrikhande R.N.,Lectr. | Bairagi V.K.,AISSMs IOIT
Proceedings - 2014 IEEE International Conference on Advances in Communication and Computing Technologies, ICACACT 2014 | Year: 2014

Lossless data compression has been suggested for many space science exploration mission applications either to increase the science return or to reduce the requirement for onboard memory, station contact time, and data archival volume. A Lossless compression technique guarantees full reconstruction of the original data without incurring any distortion in the process. In this study we evaluate the performance of several lossless grayscale image compression algorithms like CALIC. ©2014 IEEE.


Avhad K.,GHRCOEM | Limkar S.,AISSMS IOIT | Kulkarni A.,GHRCOEM
Advances in Intelligent Systems and Computing | Year: 2014

Opportunistic networks are wireless networks that have sporadic network connectivity, thus rendering the existence of instantaneous end-to-end paths from a source to a destination difficult. Hence, in such networks, message delivery relies heavily on the store-and-forward paradigm to route messages. However, limited knowledge of the contact times between the nodes poses a big challenge to effective forwarding of messages. This paper describes probability and priority based approach for routing in opportunistic networks, rendering traditional routing protocols unable to deliver messages between hosts. This approach minimizes the copies of the messages in the network and ensures successful delivery of messages. © Springer International Publishing Switzerland 2014.


Sayed A.,SAOE | Sardeshmukh M.,SAOE | Limkar S.,AISSMS IOIT
Advances in Intelligent Systems and Computing | Year: 2014

There are various Iris recognition and identification schemes known to produce exceptional results with very less errors and at times no errors at all but are patented. Many prominent researchers have given their schemes for either recognition of an Iris from an image and then identifying it from a set of available database so as to know who it belongs to. The Gabor filter is a preferred algorithm for feature extraction of Iris image but it has certain limitations, hence Principal Component Analysis (PCA) is used to overcome the limitations of the Gabor filter and provide a solution which achieves better results which are encouraging and provide a better solution to Gabor filters for Off Gaze images. © Springer International Publishing Switzerland 2014.


Kumbhar H.,SVPMs COE | Kulkarni R.,WIT | Limkar S.,AISSMS IOIT
Advances in Intelligent Systems and Computing | Year: 2014

In recent years web applications allows users to interact or collaborate with each other in a social media dialogue, in contrast to websites where very few users view the static content that was created for them. A rapidly growing trend in web applications is the development of user interfaces through Rich Internet Applications. Among other capabilities, RIAs offer high interactivity and native multimedia support, giving them a major advantage over legacy web systems. In this paper we propose a model which converts legacy web application into web 2.0 application. Conversion to the Web2.0 is twofold, i.e. automatic as well as manual. Our proposal focuses on the user interface adaptation of existing web 1.0 applications (Legacy) to Web 2.0 and add new RIA features with some good enhancement in legacy web system by web assessment. © Springer International Publishing Switzerland 2014.


Kadam M.,GHRCEM | Limkar S.,AISSMS IOIT
Advances in Intelligent Systems and Computing | Year: 2014

Recently the method called DMV (detecting malicious vehicle) for VANET is presented by many researchers in order to efficiently detect and isolate the malicious vehicle from the VANET in order to reduce the information leakage. This method is purely based on the concept of certificate authority, clustering, while lists monitoring etc. However as per the performances observed during the practical studies, these methods efficiently reduces the packet drops, however consumes more time for its processing and affecting other performance parameters like end to end delay, throughput and jitter which is the most vital for any kinds of wireless communications. Therefore to overcome these issues we have proposed method called Detection and Prevention of Misbehave/Malicious Vehicles D&PMV [1]. In our previous work we have observed that this method not affecting the performances of throughput, end to end delay and jitter. Therefore this method becomes efficient and more secured as compared to DMV. The architecture of both methods presented along with their algorithms for performance investigation. This paper emphasis on the discussion and analysis over both the methods in order to show that our proposed method gives much better performance as compared to existing methods. © Springer International Publishing Switzerland 2014.


Kadam M.,GHRCEM | Limkar S.,AISSMS IOIT
Advances in Intelligent Systems and Computing | Year: 2014

VANET means vehicular ad hoc network is nothing but the group of independent vehicles nodes which are moving throughout the wireless network freely. Such kind of networks are temporary as the vehicles and their positions are not fixed and hence the all the routing paths which are established in order to make the communication in between the source and destination are on demand and depends on the nodes movement into the network. The architecture is not at all needed for such kind of networks. Role of routing protocols is most important for the VANET which is used to route the data from source to destination, but they are also vulnerable to the many of the security attacks in the VANET. Due to the unprotected nature of the VANET networks routing protocols, such networks also unprotected from the malicious vehicles in the network itself. This paper presents new approach for not only the detection of malicious vehicles attack but also their prevention from the VANET. Proposed algorithm is referred as Detection and Prevention of Malicious Vehicles (D&PMV). The malicious vehicles detected using the monitoring process over the VANET, once they are detected, proposed algorithm is applied for the prevention of the same. The detection of malicious vehicles is based on DMV algorithm presented earlier. © Springer International Publishing Switzerland 2014.


Kadam M.,GHRCEM | Limkar S.,AISSMS IOIT
Advances in Intelligent Systems and Computing | Year: 2014

VANET means vehicular ad hoc network is nothing but the group of independent vehicles nodes which are moving throughout the wireless network freely. Such kind of networks are temporary as the vehicles and their positions are not fixed and hence the all the routing paths which are established in order to make the communication in between the source and destination are on demand and depends on the nodes movement into the network. The architecture is not at all needed for such kind of networks. Role of routing protocols is most important for the VANET which is used to route the data from source to destination, but they are also vulnerable to the many of the security attacks in the VANET. Due to the unprotected nature of the VANET networks routing protocols, such networks also unprotected from the malicious vehicles in the network itself. This paper presents new approach for not only the detection of malicious vehicles attack but also their prevention from the VANET. Proposed algorithm is referred as Detection and Prevention of Malicious Vehicles (D&PMV). The malicious vehicles detected using the monitoring process over the VANET, once they are detected, proposed algorithm is applied for the prevention of the same. The detection of malicious vehicles is based on DMV algorithm presented earlier.


Kumbhar H.,SVPM's COE | Limkar S.,AISSMS IOIT | Kulkarni R.,WIT
Advances in Intelligent Systems and Computing | Year: 2016

“Sa_Re_Ga_Ma_Pa_Da_Ni_Sa” is the soul of Indian Music. As we play these 7 nodes at different frequency, different length, and at different reparations on one or more nodes, we will listen n number of feelings, expressions, and emotions. As musician can create such number of tones manually and get famous, all world remember such tone forever. As frequency, node duration, node energy are the parameters for creating different tones for 7 nodes, we can use genetic evolutionary theory for revolutionary musical tones evolution. In this paper, we are proposing novel method for musical tone generation using machine learning algorithm with the help of Narmour Structure Analysis. © Springer India 2016.


Adhau A.S.,AISSMS IOIT | Shedge D.K.,AISSMS IOIT
Proceedings - IEEE International Conference on Information Processing, ICIP 2015 | Year: 2015

Iris accuracy stands highest by Comparing of all the biometrics. This paper proposed methods for iris edge detection, extraction of feature and matching whereas the important condition is that eye is partially closed or eye is blinked. Due to occlusion effect present in an iris image, most images are incapable to maintain false rejection and acceptance ratio. To solve this issue, various methods are highlighted in this paper. Canny operator is used in the segmentation for separate edges of iris from non-iris part. To detect iris properties, K-out-of-n and Euclidian distance methods are used. This is possible to assess the result by using database CASIA v2.0. The experimental work is performed and accuracy up to 95 % is achieved. Genuine images give out with good accuracy which can be used in security related applications. Hence system is executed for superior function of security, robustness and accuracy. © 2015 IEEE.

Loading AISSMS IOIT collaborators
Loading AISSMS IOIT collaborators