In this work a simple methodology has been proposed to categorize video sequences of moving crowd. Based on the motion pattern such crowds are classified either as structured or semi-structured or unstructured ones. Interest points detected in the first frame of the sequence are tracked over the sequence using optical flow. Thus, it requires tracking of only a subset of points in the frame. Based on the motion orientation of such tracked points descriptor is computed. By concatenating the block level histograms of motion orientation frame level feature has been computed. Thus it can well capture the localized motion patterns present in the segments of crowd. Frame level features are concatenated to represent the sequence. Finally, a neural network with multiple hidden layers has been used to classify the sequences. | Paper | Slide Deck | IEEE Explore Page |

This project was done as a part of Summer School on Computer Vision, Graphics and Image Processing, organized by the ECSU dept. of ISI, Kolkata. The project involved real time detection of hand gestures and tracking the fingers to manipulate the mouse pointers. | Presentation | Video | Project Report |

This Project explores a distributed method for common factor attack to RSA Moduli. In a memory resource constrained environment, the computational resources may not be sufficient enough to create the product tree required for batchwise-gcd of the entire dataset. As the size of the dataset increases the size of the product tree increases exponentially. Hence,we show a parallel processing method with theoretical proof to show that satisfactory level of accuracy can be achieved using this method.

In this work, we have considered low level time-domain and spectral features extracted from the music signal and applied unsupervised clustering algorithm like K-means and Agglomerative Cluestering to cluster the songs into broadly four emotion classes and then use classification for testing accuracy of the predicted classes.

Submitted and under review in Special issue for Deep Learning in Music for Neural Computing and Applications

My project was design and development of mechanism to enhance the Tizen Web Ecosystem for Samsung mobile devices. Enhancement was done by importing Google Web Store store apps to the Tizen ecosystem. It involved json and manifest parsing along-with settings, privilege and certificate management for both the systems for fully functional compliance from user perspective. Achieved an accuracy of about 90%.