Frattini Luca

Anymal_Navigation. Extend the navigation capabilities of Anymal-C in IsaacLab using Deep Reinforcement Learning

Anymal_Locomotion_Position. An Isaac Lab implementation of end-to-end position-based locomotion, training legged robots to navigate to fixed world targets using custom reward gating and curriculum learning, inspired by Rudin et al. (2022).

Parkour_Learning. End-to-end vision-based parkour policy in Isaac Lab. This project leverages Isaac Lab's manager-based architecture to train legged robots to conquer diverse obstacles using a unified neural network policy

RL-playground. Here I intend to practice and explore the realm of (deep) Reinforcement Learning, focusing primarily on the algorithmic side of things.