Aron Samuel Georgekutty

• Pefected an RTK-GNSS potato planting system with centimeter-level precision. • Embedded programming of an RC-driven Rover for the TuberScan project, with a stereoscopic depth camera and GPR for data capture. • Design and maintenance of gas sensors for early detection of tuber-based diseases, avoiding food wastage. • Conceptualized Ground Penetrating Radars as a non-intrusive approach for the detection and volumetric analysis of underground crops to maximize the yield, by using ML models on the radargrams.

• Developed free and open-source Machine Learning tools to enhance productivity and broaden accessibility in the government sector. • Created a web-based App for database creation with PHP and Flask, and made a unique dataset of handwritten word images and labels for training a deep learning model for an OCR tool in a regional language. • Guided and mentored external research interns with their Master’s thesis projects in Python, Web Scraping, and Machine Learning.

• Tackled a challenging Gesture Recognition problem on RGB-D video samples from the ChaLearn LAP dataset, released by Microsoft to improve the performance of their Kinect devices. • Comprehensive literature review and implementation of the latest deep learning frameworks in Gesture Recognition on videos, along with their performance evaluation and comparison. • Gained hands-on experience with GPU acceleration using CUDA