EGOR MATSUEV

- implemented real-time 3D and 2D object detection and visualisation in C++ which increased the system's safety of autonomous shunting locomotive - wrote automation scripts - composed dataset annotation instruction - collected, augmented, and cleaned dataset - trained model and validated results, optimised parameters - implemented plane segmentation algorithm, unit tests, refactored code in C++ - developed integration tests for 5 services in Python

- searched, bought and improved mobile robots at the Faculty of Control Systems and Robotics - ported novel color detection algorithm into robot in C++ which allowed navigation with lights off in a room (https://github.com/Laboratory-of-Embodied-Intelligence/Software/commit/3319b1114f951b265e991ee02ac6edb47fcb40c8) - researched reinforcement learning, V2X - presented results at conferences (https://kmu.itmo.ru/digests/article/3116, http://pps.ifmo.ru/docs/archive/2020v2.pdf)

- took the 6th place in international Robocup competition in the 5-person team (https://www.youtube.com/watch?v=zfFbWxl9F68, https://drive.google.com/file/d/1HQHefwEQXX3HwT0FROLisY1_R20DY-KX/view?usp=sharing) - presented robots at 4 festivals - solved kinematics and improved computer vision (https://github.com/Robator/red_manipulation_step/commit/b18d6535928a0f8a0380d6592598c1fb43087155) - replaced robot's computer, created and installed new gripper