PIRLNav: Pretraining with Imitation and RL Finetuning for ObjectNav

Open in new window