Adrian-Paul Botezatu

Gheorghe Asachi Technical University of Iași

Papers

Total Citations

H-Index

About

Adrian-Paul Botezatu is a researcher advancing the field of robotic visual servoing through deep learning. His primary focus lies in developing intelligent control systems that allow robots to navigate and manipulate their environment using visual feedback. Botezatu’s core contribution is the creation of a hybrid deep learning framework for eye-in-hand robotic systems, where a camera is mounted directly on the robot’s gripper. His innovative approach employs an early fusion technique, combining real and synthetic images to train a modified ResNet-18 architecture. This method directly computes the necessary velocities for pose alignment, bypassing the traditional, error-prone steps of feature detection and camera calibration. His work, detailed in papers like "Hybrid Deep Learning Framework for Eye-in-Hand Visual Control Systems" (2025) and "Enhancing Visual Feedback Control through Early Fusion Deep Learning" (2023), has garnered early citations, signaling its relevance to the robotics community. By tackling the long-standing difficulties of visual servoing, Botezatu is paving the way for more robust and efficient robotic control, with potential applications in manufacturing, automation, and autonomous systems.