Problem Space Transformations for Out-of-Distribution Generalisation in Behavioural Cloning
This work addresses inefficiencies in robotic manipulation algorithms for scenarios with limited demonstrations, though it appears incremental by applying known transformations to enhance existing methods.
The paper tackles the challenge of out-of-distribution (OOD) generalization in behavioral cloning for robotic manipulation, where finite datasets fail to cover the state space, by leveraging pose equivariance and locality properties to transform the problem space, resulting in improved OOD performance for policies using MLP-based or diffusion-based action prediction.
The combination of behavioural cloning and neural networks has driven significant progress in robotic manipulation. As these algorithms may require a large number of demonstrations for each task of interest, they remain fundamentally inefficient in complex scenarios, in which finite datasets can hardly cover the state space. One of the remaining challenges is thus out-of-distribution (OOD) generalisation, i.e. the ability to predict correct actions for states with a low likelihood with respect to the state occupancy induced by the dataset. This issue is aggravated when the system to control is treated as a black-box, ignoring its physical properties. This work characterises widespread properties of robotic manipulation, specifically pose equivariance and locality. We investigate the effect of the choice of problem space on OOD performance of BC policies and how transformations arising from characteristic properties of manipulation could be employed for its improvement. We empirically demonstrate that these transformations allow behaviour cloning policies, using either standard MLP-based one-step action prediction or diffusion-based action-sequence prediction, to generalise better to OOD problem instances.