In shared autonomy (SA), accurate user intent prediction is crucial for good robot assistance and avoiding user–robot conflicts. Prior works have relied on passive observation of joystick inputs to predict user intent, which works when the goals are clearly separated or when a common policy exists for multiple goals. However, they may not work well when grasping objects to perform daily activities, as there are multiple ways to grasp the same object. We demonstrate the need for active information-gathering in such cases and show how this can be done in a principled manner by formulating SA as a discrete action partially observable Markov decision process (POMDP), reasoning over high-level actions. One of our insights is that apart from having explicit information-gathering actions and goal-oriented actions, it is important to have actions that move toward a distribution of goals and provide no assistance in the POMDP action space. Compared with a method with no active information-gathering, our method performs tasks faster, requires less user input, and decreases opposing actions, especially for more complex objects, getting higher ratings and preference in our user study.

Robust object grasping under uncertainty is an essential capability of service robots. Many existing approaches rely on far-field sensors, such as cameras, to compute a grasp pose and perform open-loop grasp after placing gripper under the pose. This often fails as a result of sensing or environment uncertainty. This paper presents a principled, general and efficient approach to adaptive grasping, using both tactile and visual sensing as feedback. We first model adaptive grasping as a partially observable Markov decision process (POMDP), which handles uncertainty naturally. We solve the POMDP for sampled objects from a set, in order to generate data for learning. Finally, we train a grasp policy, represented as a deep recurrent neural network (RNN), in simulation through imitation learning. By combining model-based POMDP planning and imitation learning, the proposed approach achieves robustness under uncertainty, generalization over many objects, and fast execution. In particular, we show that modeling only a small sample of objects enables us to learn a robust strategy to grasp previously unseen objects of varying shapes and recover from failure over multiple steps. Experiments on the G3DB object dataset in simulation and a smaller object set with a real robot indicate promising results.

This paper presents experiments on the automatic recognition of roles in meetings. The proposed approach combines two sources of information: the lexical choices made by people playing different roles on one hand, and the Social Networks describing the interactions between the meeting participants on the other hand. Both sources lead to role recognition results significantly higher than chance when used separately, but the best results are obtained with their combination. Preliminary experiments obtained over a corpus of 138 meeting recordings (over 45 hours of material) show that around 70% of the time is labeled correctly in terms of role.