Jul 2022 – Aug 2024 · 2 yrs 1 mo

Aug 2021 – Jul 2022 · 11 mos

Jul 2020 – Jul 2021 · 1 yr

• 1) Human Emotion Recognition using Multimodalities || OpenCV, Tenserflow
• Proposed algorithm to reduce the features in video modality.
• Impact:- Maintain the F-Score and reduce validation time by a factor of 88.89%.
• Explored performance with ML (Bagging and Boosting) and DNN algorithms (CNN, biLSTM)
• 2) Adversarial Attacks and Defences in Malware Analysis on Android || Reinforcement Learning
• Robust Android Malware Detection System Against Adversarial Attacks Using Q-Learning.
• Maximising fooling rate for a single policy attack and multiple policy attack.
• Adversarial defence strategy which reduces the average fooling rate.

Jan 2020 – Jun 2020 · 5 mos

• Dynamic Malware Analysis on Android Platform with System Calls || Machine Learning
• Developed python framework to automate system calls extraction from android kernel.
• Proposed feature reduction algorithm without loss of accuracy ~94%.
• Impact : Reduced training and validation time by 99%.
• Performance evaluation with different ML/DNN algorithm.

Aug 2019 – Dec 2019 · 4 mos

• 32/64-bit MIPS multicycle processor design || Verilog, Xilinx
• Designed and simulated 32/64-bit MIPS multi cycle processor in Verilog.
• Designed instruction memory and register file as part of MIPS architecture.
• Designed data path and control circuitry for R, I and J type instruction.

Jun 2015 – Oct 2018 · 3 yrs 4 mos · Pune, Maharashtra, India

• Development of Tool for Display Graphics Devices.
• Development of dev build portal in Java.
• Developed scripts to traces driver "crash logs" in Python.
• Worked on Nvidia driver qulaity and performance check in virtual desktop and Hypervisor.
• Worked on deploying nvidia vGPUs and VMs in data center across different servers.
• Worked on debugging tool like "Windbg".
• Automate task execution with shell scripting.
• Creating documents and knowledgebase for team.