Rahul Nagshi

ABSTRACTWorking memory, being the capacity to hold the information about a stimulus after it has been removed, allows organisms to perform complicated actions in a constantly varying environment. There is a predominant body of neuroscientific literature, both experimental and computational, that posits enhanced persistent spiking activity as one of the neural underpinnings of working memory. Both network mechanisms as well as cell-intrinsic mechanisms (which are not necessarily opposing mechanisms) have bee

Master Thesis & Lab Rotation (Mandatory Study Internship) • Reviewing semantic segmentation models and working on enhancing the performance of one of the deep learning models for road network extraction from satellite imagery • Comparing image segmentation deep learning models for the extraction of road networks from satellite imagery • Exploring image segmentation deep learning models for road network extraction from satellite imagery

• Developing software test framework based on Google Test using C++ for an map updation software library • Proposing and designing a UI for the map updation software framework

Software Developer • C# based UI software for hardware diagnostics of vehicle simulator inputs & outputs R&D Engineer • Three Phase Smart Energy Meters Circuit design, firmware architecture design, and implementation in C on Stellaris Cortex M3 microcontroller • Aircraft Maintenance Simulator > ATMEGA32 microntroller based aircraft panel design and integration with software > Diagonostics Software for Vehicle Simulator

- Machine Learning (Supervised and Unsupervised) - Deep Learning (master thesis) - Models of Neural System - Models of Higher Brain Function