Kunal jani

Research Intern

DA-IICT•  January 2020 - May 2020

A comparison between two face recognition algorithms, namely, Principal Component based face recognition algorithm and the Discrete Cosine Transform based face recognition algorithm, comparing accuracy and execution time. Gaussian noise, quantization noise and salt and pepper noise were taken to analyze the performance in terms of accuracy and the ROC curves representing the variation of the false positive rate with the true positive rate. MPI was used for parallelization so that the execution time reduced for both algorithms.

Research Intern

DA-IICT•  May 2019 - July 2019

Principal Component Analysis of Dimensionality Reduction was used to compare a new face image with a set of face images in a data store. When the face recognition system was parallelized, a super linear speedup was observed without the accuracy of the being affected. The research paper written for this project was accepted by HiPC student research symposium.