Carlos Daniel de Sousa Bezerra
Papers
4
Total Citations
16
H-Index
2
About
Carlos Daniel de Sousa Bezerra is a rising researcher at the forefront of autonomous mobile robotics, specializing in the integration of deep reinforcement learning, sensor fusion, and efficient data transmission. His work centers on enabling robots to navigate complex, dynamic environments safely and intelligently. Bezerra’s major contributions include pioneering a Double Deep Q-Network (DDQN) approach that fuses LiDAR and camera data with people-detection-based collision avoidance, a method that has garnered 9 citations for its practical safety improvements. He further advanced the field by hybridizing Deep Q-Networks with an Extended Kalman Filter (EKF-DQN), accelerating learning curves and enhancing reward optimization in real-time navigation. Demonstrating a keen eye for real-world deployment, Bezerra also developed a novel system using autoencoders to compress LiDAR data for efficient transmission over low-bandwidth LoRa networks within the Robot Operating System (ROS). His work on training deep reinforcement learning agents in virtual scenarios for application on real robots underscores his commitment to bridging simulation and reality. With a growing citation count and a focus on scalable, secure autonomy, Bezerra is establishing himself as a key innovator in next-generation robotic navigation.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4