Research/Publications

Trajectory prediction for robots

2021-2022

I researched how to improve the trajectory prediction algorithm for RoboCup’s Small Size League (SSL) Robots, under the supervision of professor Marcos Maximo. I introduced an encode-decoder sequence-to-sequence neural network with attention to forecast the trajectory. Its performance supassed that of a Kalman predictor.

Steuernagel, L., Maximo, M.R.O.A. (2023). Trajectory Prediction for SSL Robots Using Seq2seq Neural Networks. In: Eguchi, A., Lau, N., Paetzel-Prüsmann, M., Wanichanon, T. (eds) RoboCup 2022:. RoboCup 2022. Lecture Notes in Computer Science(), vol 13561. Springer, Cham. Available on Springer

Computer vision for robots

2018-2019

As an undergraduate researcher, I worked under the supervision of prefessor Marcos Maximo on a vision algorithm to detect a soccer ball and the soccer field’s goalposts for RoboCup’s Humanoid KidSize robots. I derived a simpler arquitechture from the You Only Look Once (YOLO) neural network to detect objects and run in real time on the robot’s embedded computer.

L. Steuernagel, M. R. O. A. Maximo, L. A. Pereira and C. A. A. Sanches, “Convolutional Neural Network with Inception-like Module for Ball and Goalpost Detection in a Small Humanoid Soccer Robot”, 2020 Latin American Robotics Symposium (LARS), 2020 Brazilian Symposium on Robotics (SBR) and 2020 Workshop on Robotics in Education (WRE), 2020, pp. 1-6, doi: 10.1109/LARS/SBR/WRE51543.2020.9307038. Available on IEEE Xplore.