kshama ramesh

Domain: Deep Learning - Autonomous Driving Topic: Neural Networks Runtime Inference Optimization for Autonomous Driving and ADAS - Implementation and Optimization of the real-time night to day image translation algorithm - Co-evolutionary pruning based approach to reduce redundant filters in GANs - TensorRT custom plugins implementation - Applied methods: Python, PyTorch, Docker, ROS, HPC, GPU, TensoRT

- Lane Segmentation based on state-of-the-art approach deep neural network using aerial images - Hyperparameter optimization for improving the performance of different models - Designed visualization dashboard that helped in project overview and annotation - Performance evaluation based on IoU, F1 -score - Applied methods: Python, TensorFlow, Docker, Keras, GCP, GDAL, QGIS, NumPy, Pandas, Hadoop

Application specific software development for frontend and backend web applications. Software development language: Java Web Application technologies: CSS, HTML5, Angular JS, SQL Tools: Eclipse IDE

-Computer Vision -Real Time Systems -Software Platforms for Automotive Systems - Including CAN & FLEXRAY protocols -Digital Signal Processing -Hardware and Software Co-Design for Embedded Systems -Numerical Simulation using MATLAB