College of Engineering, Pune

About: College of Engineering, Pune is a based out in . It is known for research contribution in the topics: Sliding mode control & Control theory. The organization has 4264 authors who have published 3492 publications receiving 19371 citations. The organization is also known as: COEP.

MOORA and TOPSIS based selection of input parameter in solar powered absorption refrigeration system

Abstract: Absorption cooling using energy media like low-grade solar thermal energy and many other sources is of utmost importance for energy needs. To enhance the coefficient of performance of system, an ex...

Fuzzy stroke analysis of Devnagari handwritten characters

Abstract: Devnagari script is a major script of India widely used for various languages. In this work, we propose a fuzzy stroke-based technique for analyzing handwritten Devnagari characters. After preprocessing, the character is segmented in strokes using our thinning and segmentation algorithm. We propose Average Compressed Direction Codes (ACDC) for shape description of segmented strokes. The strokes are classified as left curve, right curve, horizontal stroke, vertical stroke and slanted lines etc. We assign fuzzy weight to the strokes according to their circularity to find similarity between over segmented strokes and model strokes. The character is divided into nine zones and the occurrences of strokes in each zone and combinations of zones are found to contribute to Zonal Stroke Frequency (ZSF) and Regional Stroke Frequency (RSF) respectively. The classification space is partitioned on the basis of number of strokes, Zonal Stroke Frequency and Regional Stroke Frequency. The knowledge of script grammar is applied to classify characters using features like ACDC based stroke shape, relative strength, circularity and relative area. Euclidean distance classifier is applied for unordered stroke matching. The system tolerates slant of about 10° left and right and a skew of 5° up and down. The system proves to be fast and efficient with regard to space and time and gives high discrimination between similar characters and gives a recognition accuracy of 92.8%.

Hybridisation of Mel Frequency Cepstral Coefficient and Higher Order Spectral Features for Musical Instruments Classification

Abstract: This paper presents the classification of musical instruments using Mel Frequency Cepstral Coefficients (MFCC) and Higher Order Spectral features. MFCC, cepstral, temporal, spectral, and timbral features have been widely used in the task of musical instrument classification. As music sound signal is generated using non-linear dynamics, non-linearity and non-Gaussianity of the musical instruments are important features which have not been considered in the past. In this paper, hybridisation of MFCC and Higher Order Spectral (HOS) based features have been used in the task of musical instrument classification. HOS-based features have been used to provide instrument specific information such as non-Gaussianity and non-linearity of the musical instruments. The extracted features have been presented to Counter Propagation Neural Network (CPNN) to identify the instruments and their family. For experimentation, isolated sounds of 19 musical instruments have been used from McGill University Master Sample (MUMS) sound database. The proposed features show the significant improvement in the classification accuracy of the system.

Convolutional neural networks for text categorization with latent semantic analysis

Abstract: The recent emphasis on intelligent systems has increased the focus on categorization techniques as it is an important step in information retrieval and natural language processing. The text categorization is largely achieved using machine learning techniques. In most of the approaches one-hot encoding or pre-trained word embedding such as word2vec or glove vectors are used. This study explores the feature vectors based encoding using Latent Semantic Analysis (LSA) technique along with the Convolutional Neural Network (CNN) being used as a classifier. It was found that applying LSA followed by CNN for text classification offers better accuracy than the conventional methods of CNN with other approaches. This research, thus, highlights the importance of Latent Semantic Analysis technique coupled with convolutional neural networks for text classification.

Optimization of Process Parameters to Study the Influence of the Friction in Tube Hydroforming

Abstract: Tube hydroforming (THF) is a well-known metal forming technology. This technology enables the manufacturing of a variety of intricate shape parts used in automobile industry. Tribology plays an important role in THF, required in the automobile industry. THF process is influenced by many process parameters. Friction between outer surface of the tube and the inner surface of the die is significant and influences the process parameters and quality of components. The aim of the proposed work is to optimize the different process parameters which influence the coefficient of friction in the THF using mathematical model based upon the tube upsetting method. Influence of friction on process parameters, mainly inner pressure and wall thickness, is analyzed and optimized. The proposed mathematical model is verified by comparison of coefficient of friction with original values for Steel35NBK and AlMgSi materials. COF (μ) decreases from 0.15 to 0.0289 for Steel35NBK and from 0.1 to 0.0136 for AlMgSi after optimization of initial tube thickness, S 0 = 3.5 mm and pressure p i = 142.9554 MPa for Steel35NBK and pressure p i = 143.5730 MPa for AlMgSi.