Working on developing end-to-end deep learning solutions for various computer vision problems.

• Worked on improving the accuracy of Convolutional Neural Network (CNN) based model for background removal.• Implemented color & noise augmentation to improve the robustness of the network.• Applied weight-based pruning to achieve better perf with no quality degradation.

• Designed and developed key components of Tegra mobile camera software driver across multiple operating systems and products - NVIDIA’s Jetson TX1, Jetson TK1, SHIELD, HTC Google Nexus 9.

The project involved modification and up-gradation of a generic THS (Test Harness Setup) which could simulate upper and lower layers for a mobile communication protocol layer.

I worked as a research intern during the period of June-July 2011, on the project "Video segmentation using neural network and distributed genetic algorithm".

I worked as a vocational trainee at BEL during the period of May-June 2011. My project title was "Study of Radar Transmitter(Rohini Radar)". The training included familiarization with all areas of manufacturing and testing of radar parts in BEL Ghaziabad which are cumulative for manufacturing of a radar transmitter. Followed by study of "rohini radar" transmitter.