VARUN KHARE

Automation of Computer Vision Tasks through cutting edge deep learning techniques, to save manual labor and operational costs of Clients. Implementation and Optimization of state-of-the-art Convolution Neural Networks ( CNNs ) for Object Detection in real time. Establishment of Client Server Architecture for efficient Communication and large data processing. Implementation of Automated Number Plate Recognition (ANPR) in real time to automatically recognize License plates in Gulf Cooperation Council (GCC) region in real time through YOLOv7 algorithm, achieving overall Accuracy of 92 %. Implementation of Automated Vehicle Type Classification on real and synthetic data, in real time through Transfer Learning technique, achieving overall Accuracy of 95 %. Containerization through Docker for automatic deployment and portability.

Mentored 150+ students in AI and Machine Learning classes. Mentored 150 + students for Practical labs in C, Python and Data Science. Supervising and upskilling 15+ B.Tech. final year students in their Machine learning and data science projects of Semantic Segmentation of Image, Classification, Customer Segmentation, Outlier detection and Time Series analysis.