Demo of Self-Supervised Gait Learning using Bursting CPG and Obstacle Avoidance using DSM for Robots

Описание к видео Demo of Self-Supervised Gait Learning using Bursting CPG and Obstacle Avoidance using DSM for Robots

In this demo video, we show the Petoi quadruped bot learn to walk and crawl using Spike-Time Dependent Plasticity (STDP) learning. A tunable Central Pattern Generator (CPG) provides rhythmic bursting spikes that govern the timing of motor commands sent to the bot. Later the bot uses the learned gaits to navigate and avoid obstacles autonomously using a Dynamical State Machine (DSM).

This video was created by ENIGMA lab team members at the University of Pittsburgh for the IEEE Transactions on Cognitive and Developmental Systems (TCDS) Journal.
https://www.engineering.pitt.edu/subs...

Комментарии

Информация по комментариям в разработке

Похожие видео