Marimuthu Selvam

Passionate Robotics Sofware Engineer with 7+ years of experience in SLAM, computer vision, and AI-driven robotics. Aspiring to pursue cutting-edge research in robot perception, deep learning, and autonomous systems to advance AI-powered robotics.

1. Dynamic Obstacle Detection for SLAM
Developed unsupervised learning models and probabilistic filters for real-time obstacle detection, improving SLAM accuracy in dynamic environments.
2. Optimization-Based SLAM Map Merging
Applied optimization algorithms to merge multiple SLAM maps, enhancing spatial accuracy and robustness in large-scale environments.
3. SLAM Optimization for Dynamic Environments
Designed a sensor fusion-based SLAM system integrating probabilistic models to improve map reliability and adaptability in changing environments.
4. AI-Powered Safe Trajectory Planning
Developed AI-driven trajectory planning algorithms for dynamic obstacle avoidance, improving robot navigation safety.
5. Hand Gesture Recognition Using CNN
Implemented a CNN-based real-time hand gesture recognition system, enabling intuitive human-robot interaction for robot control.
6. Bridging ROS2 with LG Framework
Designed and implemented a network bridge using ZeroMQ, optimizing communication and performance between ROS2 and LG robotic frameworks.