Work on the development of a singular computer vision solution to parse wide range of visual data (PDFs, Excels, TXTs, etc.)

• Developed pose estimation and correction model to capture,evaluate and provide feedback on user performance• Designed a 3D pose differentiation algorithm that accounts for time-dilation using Dynamic Time Wrapping• Developed a generalized scoring mechanism and a live feedback mechanism, to ensure forward compatibility.

• Collected data about the existing ETP and calculated its efficacy and compliance with national standards• Identified points of failures and ranked them on their latent potential in improving the system performance• Recommended key system modifications to overcome the flaws and reported their feasibility and ROI estimates

• Analyzed numerous deep Convolutional Neural Network (CNN) architectures for reliable Image Segmentation• Implemented a face-liveliness detection system as a measure against photographic and video graphic spoofing• Achieved 86% accuracy by optimizing the existing face-recognition system on masked or partially obstructed faces