@inproceedings{Valassakis:2021:10.1109/IROS45743.2020.9341617,
author = {Valassakis, P and Ding, Z and Johns, E},
doi = {10.1109/IROS45743.2020.9341617},
publisher = {IEEE},
title = {Crossing the gap:  a deep dive into zero-shot sim-to-real transfer for dynamics},
url = {http://dx.doi.org/10.1109/IROS45743.2020.9341617},
year = {2021}
}

Download

TY  - CPAPER
AB  - Zero-shot  sim-to-real  transfer  of  tasks with complex dynamics  is  a  highly challenging and unsolved problem. A number of solutions have been proposed in recent years, but we  have found that  many works do  not  present  a  thorough evaluation in the   real world,   or   underplay the   significant engineering effort and task-specific fine tuning that is required to  achieve  the  published results.  In  this paper,  we  dive  deeper into  the  sim-to-real  transfer  challenge,  investigate  why  this  issuch a difficult problem, and present objective evaluations of anumber of transfer methods across a range of real-world tasks.Surprisingly,  we  found  that  a  method  which  simply  injects random  forces  into  the  simulation  performs  just  as  well  as more  complex  methods,  such  as  those  which  randomise  the simulator's dynamics parameters
AU  - Valassakis,P
AU  - Ding,Z
AU  - Johns,E
DO  - 10.1109/IROS45743.2020.9341617
PB  - IEEE
PY  - 2021///
TI  - Crossing the gap:  a deep dive into zero-shot sim-to-real transfer for dynamics
UR  - http://dx.doi.org/10.1109/IROS45743.2020.9341617
UR  - http://hdl.handle.net/10044/1/83620
ER  - 

Download