College of Engineering, Pune

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

“Feature level fusion of face, palm vein and palm print modalities using Discrete Cosine Transform”

Abstract: Due to usefulness in recognition and identification biometric systems have become a major part of research. Paper proposes a multimodal biometric system using face modality combined with palm print and palm vein modality. The proposed methodology uses Local Statistical method in which pre-defined block of DCT coefficient were used to calculate standard deviation and store them as feature vector. Matching is done using distance between feature vector of testing and training data set. Results show that the Genuine Acceptance Rate (GAR) of feature level fusion is 100% which is better than, that of uni-modal systems, hence having multimodality is advantageous. For testing and training database of 100 students of College of Engineering Pune.

Secure and efficient message passing in distributed systems using One-Time Pad

Abstract: Distributed Computing system is a cluster of computers connected on a network which coordinate their actions via passing messages. Since the cluster can also be accessed publicly, the security of various messaging operations is of great concern. In this paper a highly efficient and secure implementation of message passing interface (MPI) in distributed environment has been proposed and implemented. This implementation, which is named as VAN-MPICH2, integrates security measures to ensure data confidentiality using One-Time Pad (OTP) encryption technique. Since the proposed encryption implementation decreases the security overhead substantially, VAN-MPICH2 manages to provide confidentiality with very insignificant decrease in performance.

Augmenting e-commerce product recommendations by analyzing customer personality

Abstract: Customer specific personalization has become imperative for e-commerce websites, helping them to convert browsers (visitors) into buyers. The e-commerce industry predominantly uses various machine learning models for product recommendations and analyzing a customer's behavioral patterns, which play a crucial role in exposing customers to new products based on their online behavior. Psychology studies show that if customers are shown products suited to their personality type or complementing their lifestyle, the chances of them buying the said product grow considerably. By incorporating the personality of a customer in a recommendation system, can we achieve increased level of customer-personalization? The answer to this question forms the crux of this paper. With a view to ascertain a customer's personality, we obtain relevant markers from text samples along the five psychological dimensions. We then experiment with various classification models and analyze the effects of different sets of markers on the accuracy. Results demonstrate certain markers contribute more significantly to a personality trait and hence give better classification accuracies. Considering the existence of an ecommerce based conversational bot, we utilize the personality insights to develop a unique recommendation system based on order history and conversational data that the bot-application would gather over time from users.

Automatic Mood Classification Model for Indian Popular Music

Abstract: Music shares a very special relation with human emotions. We often choose to listen to a song or music which best fits our mood at that instant. A lot of research and study has been going on in the field of Music mood recognition in the recent years. We contribute to make an effort for automatic identification of mood underlying the audio songs by mining their spectral and temporal audio features. Our current work involves analysis of various classification algorithms in order to learn, train and test the model representing the moods of the audio songs. The focus is on the Indian popular music pieces and our work continues to analyze, develop and improve the algorithms to produce a system to recognize the mood category of the audio files automatically. The experimental results show a satisfactory performance of the system in recognizing the music mood by using ensemble classification tree techniques.