Start from the trot: behaviour cloning and residual policies

Training from scratch found small, fast steps that hobby servos may not follow. Two ways to start from the gait that already walks instead: copy it into the policy network and let PPO improve it, or keep the trot running and learn only a small correction on top. Both trained, measured on random robots and compared.

55 min

Esta página aún no está traducida, por eso se muestra en inglés.

Esta lección forma parte del curso de pago. La primera lección de cada curso es gratuita.

Pedir por correo24 €

Respondemos en un día con los datos de pago (transferencia o PayPal) y te enviamos los archivos por correo justo después del pago.