Developed a custom CNN architecture to classify earthquake magnitudes from seismograms. Spearheaded a feasibility study on integrating Deep Learning in Seismology in close collaboration with esteemed faculty at CDMM. Implemented a custom 3D CNN and GRU architecture for real-time earthquake detection. Transformed seismograms into Mel-Log spectrograms to enhance input quality, contributing to achieving accurate earthquake magnitude classification.

Collaborated as a Student Research Intern under the mentorship of a distinguished professor in the Department of Analytics. Developed a cutting-edge project focused on the classification of Indian Currency Banknotes utilizing Convolutional Neural Networks (CNN). Leveraged Transfer Learning techniques to evaluate and enhance the performance of multiple CNN models, achieving an outstanding overall accuracy rate of 99.05% (f-score).