R. Ramya
@ksrct.ac.in
Assist.Prof
K.S.Rangasamy College of Technology
EDUCATION
B.E, M.E (
RESEARCH INTERESTS
Machine Learning , Image Processing, Agriculture
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
R.S. Ramya, Sandhya K, K.R. Venugopal, S.S. Iyengar, and L.M. Patnaik
IEEE
Mining opinions from online reviews is an essential step in obtaining the overall sentiment of a product. Deep learning procedure is applied over various fields. User ratings are huge for recommender structures since they consolidate various kinds of energetic information that may influence the exactness of the suggestion. In this work, a deep learning model is utilized to process the user remarks and to create a potential user rating for user comments is proposed. To start with, the system uses sentiments to create a feature vector as the input nodes. Further, the framework tools reduce the noise in the dataset to recover the classification of information mining. To finish, Deep Belief Network (DBN) and sentiment analysis reaches data learning for the approvals.
R. Ramya and S. Moorthi
Springer Science and Business Media LLC
R. Ramya, K. Selvi, K. Murali, and Gajana Penchalaiah
Springer Singapore
M. Karthick, K. S. Praveen Kumar, Pranish Nikile C.V., R. Ramya, and S. R. Mohanrajan
IEEE
This paper discusses the procedure and steps followed to design and develop an electric hover-board. The design for the mechanical chassis, an approach to calculate the required motor power rating depending on the load, the possibilities of using different motors, methods to control the rotation of the BLDC motor and closed speed control algorithms for BLDC motor are all presented. Design and development of motor driver required to build a hoverboard is also discussed. The development of motor driver includes design of inverter, DC-DC converter, level shifter and fabrication of the PCB. On completion of fabrication, the motor is tested for open loop control and closed loop control using hysteresis.
R. Ramya, P. Kumar, K. Sivanandam, and M. Babykala
IEEE
Fruit disease detection is vital at early stage since it will affect the agricultural field. In this paper, mainly consider the detection and analysis of fruit infections which is available in the plant areas and storage of data about the agricultural filed and details of farmers in database and recovering the data using Cloud computing. There are more fruit diseases which occur due to the surrounding conditions, mineral levels, insects in the farm area and other factors. The detected data from the plant area is determined by image processing and stored in the database.
Seema Singh, R. Ramya, V. Sushma, S.R. Roshini, and R. Pavithra
IEEE
Facial recognition is a non-invasive method of biometric authentication and useful for numerous applications. The real time implementation of the algorithm with adequate accuracy is required, with hardware timing into consideration. This paper deals with the implementation of machine learning algorithm for real time facial image recognition. Two dominant methods out of many facial recognition methods are discussed, simulated and implemented using Raspberry Pi. A rigorous comparative analysis is presented considering various limitations which may be the case required for innumerable application which utilize facial recognition. The drawbacks and different use cases of each method is highlighted. The facial recognition software uses algorithms to compare a digital image captured through a camera, to the stored face print so as to authenticate a person's identity. The Haar-Cascade method was one of the first methods developed for facial recognition. The HOG (Histogram of Oriented Gradients) method has worked very effectively for object recognition and thus suitable for facial recognition also. Both the methods are compared with Eigen feature-based face recognition algorithm. Various important features are experimented like speed of operation, lighting condition, frontal face profile, side profiles, distance of image, size of image etc. The facial recognition model is implemented to detect and recognize faces in real-time by means of Raspberry Pi and Pi camera for the user defined database in addition to the available databases.
R.S. Ramya, Naveen Raju, N. Sejal, K.R. Venugopal, S.S. Iyengar, and L.M. Patnaik
IEEE
A user query facet is a collection of items that summarizes the content covered by a query. In general, the most significant information of a user query is present in the top retrieved document that are in the form of lists. In this work, we propose a framework Automatic Extraction of Facets for User Queries [AEFUQ] that extract the user query facets automatically by grouping the list based on three categories namely HTML tags, free text patterns and repeat regions. Grouping of the list is based on domain sites present in the list. We observe that some of the lists are not relevant for extracting the facets. In order to prune these lists, the importance of each item in the lists that are present in the group G is evaluated and Cosine Similarity (CS) between two items is calculated. Further, based on CS score obtained, High Quality Clustering (HQC) algorithm is proposed to cluster the items that has the most number of point in each iteration to obtain more number of facets. Finally, the top most items from each cluster are provided as the best facets for the user query. Experiments are conducted on User Q and Random Q dataset. It is observed that the proposed method AEFUQ outperforms by providing a large number of useful query facets compared to QDMiner method [1].
R S Ramya, M Darshan, N Sejal, K R Venugopal, S S Iyengar, and L M Patnaik
IEEE
Owing to the steep increase in the Internet population, the content over the web is increasing exponentially so as Community Question Answer (CQA) have acquired very huge amount of questions and answers. In this article, a machine learning algorithms are utilized for Question Classification (QC) and Answer Classification (AC). We identify the category of the question posted and further map with the corresponding question. Similarly for the answers posted by the multiple user will be processed for category mapping. Here the result shows the effective classifier that can be chosen to perform the mapping task for both Question classifier as well as answer classifier. Here the results shows that, for Question Classification (QA), Linear Support Vector Classification (LSVC) is found to be best classifier and Multinomial Logistic Regression (MLR) is most suitable for Answer Classification (AC). Using the probability of overall possible outcomes of a particular answer will give a best answer. Experiments results shows that our method outperforms efficiently.
R. Ramya and S. Moorthi
Springer Science and Business Media LLC
AbstractHardware multiplier circuits decide the speed and power consumption in the execution of digital signal processing algorithms. The desirable feature of reduced area and power consumption for battery-driven multimedia gadgets can be realized by replacing the power hungry multiplier circuits with approximate multiplier circuits. The approximation techniques reduce the complexity of the design and improve the energy efficiency of the circuit. This paper proposes an area and power efficient approximate unsigned integer multiplier architecture based on wordlength reduction. It is designed to meet a pre-specified error performance with improved area and power reduction compared with similar designs. It is extended further for the signed multiplier architecture. The circuit characteristics are analyzed to establish the suitability of the proposed design for low-power applications. Synthesis results show that the proposed unsigned multiplier consumes 65% less power than the exact Wallace multiplier. The area requirement of the proposed multiplier reduces by 50% compared to an exact multiplier. The multiplier is tested for image filtering to establish the efficacy of the design in multimedia applications.
Ramya R S, Nanda Kishore, Sejal D, Venugopal K R, S S Iyengar, and L M Patnaik
IEEE
Due to the rapid rise in internet population, the content over web is increasing and a large number of documents assigned by reader’s emotions have been generated through new portals. Earlier works have focused only author’s perspective, our work focuses on reader’s emotions generated by news articles. Social emotions of news articles from reader’s perspective are predicted with the help of user ratings. More specifically, we form Communities based on the ratings that are present in the news articles. Further, a Textual Relevance is computed based on the word frequency for a particular document. Experiments are conducted on the news articles and as a result, it is observed that the proposed method results in predicting reader’s emotions are much better when compared with the existing method Opinion Network Community (ONC) [1].
Ramya R and Moorthi S
Springer Science and Business Media LLC
The advancements in medical healthcare networks and bio-medical sensor technologies enabled the use of wearable and body implantable intelligent devices for healthcare monitoring. These battery-operated devices must be capable of very low power operation for ensuring long battery life and also to prevent intense radiations. The major power consuming part of these devices are the multipliers built into the digital filters for performing signal processing operations. This paper proposes a low power signed approximate multiplier architecture for bio-medical signal processing applications. The circuit characteristics and error metrics of the proposed multiplier are estimated to verify its performance advantage over other approximate multipliers. In order to validate the efficacy of the approximate multiplier in real time signal processing applications, a band pass finite impulse response filter (FIR) filter is designed using frequency response masking technique and used in the Pan Tompkins method for the extraction of QRS complex from raw ECG data. The sensitivity, positive predictivity, and detection error rate of the QRS detection method are estimated and the results show that the approximate filtering method implemented gives a comparable performance as that of exact methods.
R S Ramya, M Darshan, Naveen Raj, D Sejal, K R Venugopal, S S Iyengar, and L M Patnaik
IEEE
Due to the rapid rise in the geo-positioning technologies and location based services, millions of spatio textual objects are collected in many applications like social networks, geo location services that has been attracted many research communities. Each spatial objects is described by its spatial locations (latitude, longitude) and a set of query Terms. In this paper, we propose a framework called Efficient Batch Top k Spatial Term (Keyword) Search by Feature Redundancy (k STFR), that retrieves top k results for the batch of queries. The clusters and subclusters are constructed based on the computed range and category of the object present in the user query. Experiments are conducted on Geographic Names (GN) dataset. Experiments results shows that our proposed method outperforms Inverted Linear Quadtree (ILQ) [1] efficiently and also support batch of queries with improved response time.
R Ramya and K Padmapriya
American Scientific Publishers
M. Saravanan, A. Muthukumar, R. Ramya, K.K. Rashika, and S. Saravanan
IEEE
In water supply networks system, the major problem is water leakage. Unwanted water leakage due to leaky pipe lines and beneath the underground pipelines is almost always pertaining in drinking water supply networks. This system contains two sections, first part is leakage detection and automatically closes the solenoid valve for to prevent the over leakage of water and send SMS to the corporation using GSM module according to sensor information. By using GPS location to detect where the leakage takes place. The second part is that to fill the water tank by using android application.
S. Sreejith, R. Ramya, R. Roja, and A. Sanjay Kumar
IEEE
This paper entitled "Smart Bin for Waste Management System" plays a vital role in the waste management system. A healthy domain is essential to a solid and cheerful environment. Clean and hygienic environments are a key need in human habitable environments. Smart bin is to develop a gainful and dynamic waste administration framework. In public places, dustbins are being flooded just as the waste spills out bringing about contamination. This likewise expands number of infections as huge number of bugs to breed on it. In this a smart bin is developed to monitor the level of waste, automatic disposing of waste and rain detection system. The outcome demonstrated that the detecting framework is effective and savvy and can be utilized to robotize any solid waste bin management process.
R. Ramya and S. Moorthi
IEEE
This paper proposes accuracy configurable multi-precision multiplier architecture suitable for signal processing applications. The proposed multiplier can be operated in approximate mode as well as in full-precision mode with variable precision capabilities. The fundamental processing element (PE) is an N/4-bit unsigned multiplier. Sixteen such multipliers are arranged in such a way that N/4, N/2, 3N/4 and N-bit multiplications can be performed in a recursive fashion by combining many or all of the sixteen N/4-bit multipliers depending on predefined modes of operation controlled by the wordlength of input operands. Simulation results show that the proposed design gives significant power savings than the exact multiplier. The proposed architecture can be configured to adapt different levels of precision and well suited for coarse grain reconfigurable architectures.
R. S. Ramya, Ganesh Singh T., D. Sejal, K. R. Venugopal, S. S. Iyengar, and L. M. Patnaik
ACM Press
In recent years, the availability of digital documents over web is increased drastically and there is a need for effective methods to retrieve and organize the digital documents. Since data is dispersed globally and is unorganized, it is a challenging task to develop an effective methods that can generate high quality features in these documents. It is necessary to reduce the gap between users search intention and the retrieved results known as semantic gap. In this paper, Discovering Relevant Documents using Latent Dirichlet Allocation and Cosine Similarity (DRDLC) is proposed. Word similarity is computed using CS Cosine Similarity present in search results documents. LDA is applied on extracted patterns and documents. Hashing is used to extract high relevant documents efficiently. Further, term synonyms are identified using word net and the documents are re-ranked. Experiments using the model Relevance Feature Discovery (RFD) on Reuters Corpus Volume-1 (RCV-1) show that the proposed DRDLC framework results in improved performance by providing more relevant documents to the user input query.
Ramya Raghavachari and M. Sridharan
The tremendous increase in the use of portable electronic devices is due to the development in the fields of signal processing and electronic technology. These battery operated devices needs reduction in power consumption with increased performance and long battery life. Since CMOS technology scaling fast approaches its physical limit of minimum supply voltage and smaller feature size, the hardware designer has to opt for new multiplier architectures for achieving low power and high speed performance. This paper proposes an area and power efficient approximate multiplier architecture. The error metrics are estimated to verify its performance advantage over other approximate multipliers. Using Frequency Response masking approach, a 6-band non-uniform digital FIR filter bank is developed using approximate multiplier for hearing aid application. Audiogram matching is done with audiograms of two different types of hearing losses and the matching error is computed. Simulation results show that the audiogram matching error falls within +/- 4 dB range.
Ramya Raghavachari and M. Sridharan
Coarse grained reconfigurable architecture got the attention of researchers working in designing computing architectures for processing massive streaming data associated with the multimedia applications in portable entertainment and communication electronics. The algorithms for processing audio, video, and graphics are very complex in nature. These data intensive computation algorithms belong to the domain of signal processing. As the complexity of algorithms increases, a matching improvement in speed performance of the hardware becomes essential to maintain the quality of service. The observed growth of algorithmic complexity is much higher than the growth rate of integration density governed by Moore’s law. Also, the constraints on memory bandwidth in the traditional von Neumann architectures along with the slow growth in the battery capacity demands a paradigm shift in computer architecture design. Reconfigurable hardware architecture is proposed as a possible alternative in this regard. The reconfigurable architectures are designed to exploit the regular and repetitive structure of signal processing algorithms and the coarse grained processing elements are designed to match with the word level granularity of these complex algorithms. The research shows that the coarse grain reconfigurable architectures with heterogeneous processing elements are a better option for system design in DSP applications which exploit granularity matching between the algorithms and the processing hardware, and also the inherent parallelism of DSP algorithms for the realization of low power DSP systems.
Sundararaman Rajagopalan, Sivaraman Rethinam, V. Lakshmi, J. Mahalakshmi, R. Ramya, and Amirtharajan Rengarajan
IEEE
The advancements happening in the domain of information technology resulted in growing stature of rapid communication across the world. Telemedicine is one such arena which is benefitted largely because of such revolution. DICOM images are one of the important medical information carriers shared mostly through an unsecured network across the hospitals and health centers. Protection of such significant medical records against unauthorized access needs an important attention. In this work, the medical image is stored in on-chip memory of FPGA and upon authentication through an unique key received through the keyboard, the image will be displayed in a VGA monitor. In the event of detecting wrong password / key, the encryption of medical image has been carried out using Cellular Automata (CA). Authentication has been improved by hardware triggered password. The proposed multiplexed image authentication was implemented on reconfigurable hardware platform Cyclone II FPGA. Hardware utility and power dissipation have been analysed as a part of this work.
R. Ramya, G. Saravanakumar, and S. Ravi
Springer India
In recent years, wireless sensor networks (WSNs) have grown dramatically and made a great progress in many applications. But having limited life, batteries, as the power sources of wireless sensor nodes, have restricted the development and application of WSNs which often requires a very long lifespan for better performance. In order to make the WSNs prevalent in our lives, an alternative energy source is required. Environmental energy is an attractive power source, and it provides an approach to make the sensor nodes self-powered with the possibility of an almost infinite lifetime. The goal of this survey is to present a comprehensive review of the recent literature on the various possible energy harvesting technologies from ambient environment for WSNs.
N Vijayakumar and R Ramya
IEEE
In order to ensure the safe supply of the drinking water the quality needs to be monitor in real time. In this paper we present a design and development of a low cost system for real time monitoring of the water quality in IOT(internet of things). The system consist of several sensors is used to measuring physical and chemical parameters of the water. The parameters such as temperature, PH, turbidity, conductivity, dissolved oxygen of the water can be measured. The measured values from the sensors can be processed by the core controller. The raspberry PI B+ model can be used as a core controller. Finally, the sensor data can be viewed on internet using cloud computing.