Ahmad Asghar

- Creation of AI-ready image dataset for Autonomous Rail Systems. - Image similarity analysis on obstacle detection (OD) and track intrusion detection (TID) datasets through feature extraction.

- Design and implementation of a novel deep learning-based image dataset reduction and clustering approach exploiting deep image feature representations. - Development of image recommendation and discard systems. - Development, training and optimization of supervised learning networks for object detection and instance segmentation. - Created data-specific training pipelines and built fully customized custom training loops.

- Accurate behavioral modeling of closed-loop Phase Locked Loop. - Development of testbenches for Low Noise Amplifiers. - Worked on the development of an approach for the characterization of Radio Frequency amplifiers in SystemC-AMS.

- Core research areas: Surface metrology and machine vision applications - Development of algorithms for simulation and improvement of surface roughness measurement techniques using MATLAB. - Performed deformation measurements of optical and high-precision 3D objects using a confocal microscope.