Conference Papers
Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506
Browse
5 results
Search Results
Item Detection of similarity in music files using signal level analysis(Institute of Electrical and Electronics Engineers Inc., 2017) Thomas, M.; Jothish, M.; Thomas, N.; Koolagudi, S.G.; Vishnu Srinivasa Murthy, Y.V.In today's age of digital media, the collection of music files available to the general public is extremely diverse. As with any such set of data, efforts must be made to classify and categorize these files in order to facilitate easy access and searching. Songs can be classified based on attributes available in the music file's metadata such as artist, album, year of release, length, etc. However, if the similarity between two songs is to be determined, a simple comparison of metadata is not only unsatisfactory, the metadata itself might not be available. Therefore, a method of comparison independent of the availability of metadata is required. In this work, a comparison method has been proposed involving the use of musical parameters such as tempo, key and signal envelope, which are extracted from the music file through signal level analysis. Genre is also computed using a support vector machine (SVM) classifier and used to estimate the similarity between two songs. © 2016 IEEE.Item An approach to maintain attendance using image processing techniques(Institute of Electrical and Electronics Engineers Inc., 2017) Yuvaraj, C.B.; Madikeri, M.; Santhosh Kumar, V.; Vishnu Srinivasa Murthy, Y.V.; Koolagudi, S.G.Nowadays, the research is growing towards the invention of new approaches. One such most attracted application is face recognition of image processing. There are several innovative technologies have been developed to take attendance. Some prominent ones are biometric, thumb impressions, access card, and fingerprints. The method proposed in this paper is to record the attendance through image using face detection and face recognition. The proposed approach has been implemented in four steps such as face detection, labelling the detected faces, training a classifier based on labelled dataset, and face recognition. The database has been constructed with the positive images and negative images. The complete database has been divided into training and testing set and further, processed by a classifier to recognize the faces in a classroom. The final step is to take the attendance using face recognition technique in which the input image of a classroom is given, and faces of the given image will be detected along with their IDs. The frames of a video taken for a minute is taken into consideration to avoid the missed ones due to rotational issues. © 2017 IEEE.Item Beat Onset Detection in an Audio Clip of a Percussion Instrument-Mridanga(Institute of Electrical and Electronics Engineers Inc., 2018) Vishnu Swaroop, G.; Koolagudi, S.G.; Vishnu Srinivasa Murthy, Y.V.The process of automating MIR tasks is essential due to the availability of enormous number of tracks. Of these, beat onset detection is a base for the task of Tala identification which is a part of Indian Classical Music (ICM). In this paper, an effort has been made to detect the onset of a specific percussion instrument called Mridanga as it is highly used instrument in Carnatic music. The dataset has been recorded at studio by playing the Mridanga for different Talas. Further, various signal to noise ratio (SNR) values have added in the range of 40 dB - 10 dB to generalize the system for real-time applications. The features such as centroid flux, and rate of change in energy have been computed for every sub-band. Various filtering approaches have been used to optimize the process of stroke onset detection. The results are found to be appreciable and an average accuracy of 95.01% is obtained with 40dB and 89.73% is with 10dB. Copy Right © INDIACom-2018.Item The Performance of Weather Forecasting using Data Mining and Evolutionary Algorithms: A Comparative Study(Institute of Electrical and Electronics Engineers Inc., 2018) Veera Ankalu, V.; Apparao, G.; Vishnu Srinivasa Murthy, Y.V.Weather plays an important role in day-to-day life. The process of predicting weather is an emerging issue to be solved and need historical quantitative data for proper analysis. The present state of atmospheric conditions will help to analyze how weather evolves from state to state. Weather prediction is basically based upon the historical time series data. The basic Data mining operations and Numerical methods are employed to get a useful pattern from a huge volume of data set. Different testing and training scenarios are performed to obtain the accurate result. To perform these kinds of predictions we are identifying the datasets. Collection of the data sets of a particular region weather report from 1901 to 2001 with 11 attributes. The collected datasets undergo pre-processing. Then clustering operation, Curve fitting and Extrapolation methods are applied, proceeding with back propagation. The Back propagation and Extrapolation results are compared. The Best future results are predicted. Copy Right © INDIACom-2018.Item Food classification from images using convolutional neural networks(Institute of Electrical and Electronics Engineers Inc., 2017) Attokaren, D.J.; Fernandes, I.G.; Sriram, A.; Vishnu Srinivasa Murthy, Y.V.; Koolagudi, S.G.The process of identifying food items from an image is quite an interesting field with various applications. Since food monitoring plays a leading role in health-related problems, it is becoming more essential in our day-to-day lives. In this paper, an approach has been presented to classify images of food using convolutional neural networks. Unlike the traditional artificial neural networks, convolutional neural networks have the capability of estimating the score function directly from image pixels. A 2D convolution layer has been utilised which creates a convolution kernel that is convolved with the layer input to produce a tensor of outputs. There are multiple such layers, and the outputs are concatenated at parts to form the final tensor of outputs. We also use the Max-Pooling function for the data, and the features extracted from this function are used to train the network. An accuracy of 86.97% for the classes of the FOOD-101 dataset is recognised using the proposed implementation. © 2017 IEEE.