Alexandre Ricardo Soares Romariz

Papers

Total Citations

H-Index

About

Alexandre Ricardo Soares Romariz is a researcher whose work bridges computer vision and robotics, with a focus on human action recognition and bio-inspired locomotion. His most cited paper, "Human Action Recognition Based on a Two-stream Convolutional Network Classifier" (2017, 18 citations), demonstrates his contribution to video analysis, leveraging deep learning to interpret human movements from portable media—a key step toward autonomous surveillance and human-computer interaction. In robotics, Romariz has advanced quadruped and humanoid gait generation. His 2018 study on "Walking Pattern Design and Balance Control of a Quadruped Platform" (4 citations) uses Multi Objective Genetic Algorithms to optimize gaits, validated in both simulation and real hardware. Earlier work, "Gait generation for a quadruped robot using Kalman filter as optimizer" (2009, 3 citations), models leg kinematics akin to parallel manipulators, while "Neural oscillator for gait command of a humanoid robot" (2012, 2 citations) applies central pattern generators (CPGs) for rhythmic bipedal motion. Though his citation counts are modest, Romariz’s integration of neural networks, optimization algorithms, and bio-inspired control offers practical insights for robotics and vision systems, appealing to students interested in applied AI and mechanical design.