Disturbance-injected Robust Imitation Learning with Task Achievement
Hirotaka Tahara,Hikaru Sasaki,Hanbit Oh,Brendan Michael,Takamitsu Matsubara,Hirotaka Tahara,Hikaru Sasaki,Hanbit Oh,Brendan Michael,Takamitsu Matsubara
Robust imitation learning using disturbance injections overcomes issues of limited variation in demonstrations. However, these methods assume demonstrations are optimal, and that policy stabilization can be learned via simple augmentations. In real-world scenarios, demonstrations are often of diverse-quality, and disturbance injection instead learns sub-optimal policies that fail to replicate desi...
