1. What are the contributions in "Object learning through active exploration" ?
This paper addresses the problem of active object learning by a humanoid child-like robot, using a developmental approach.. The authors propose a cognitive architecture where the visual representation of the objects is built incrementally through active exploration.. The authors present the design guidelines of the cognitive architecture, its main functionalities, and they outline the cognitive process of the robot by showing how it learns to recognize objects in a human-robot interaction scenario inspired by social parenting.. The authors discuss the further benefits of their architecture, particularly how it can be improved and used to ground concepts.
read more
2. What is the primary form of communication that a baby has with his mother?
Contact and touch are also the primary form of communication that a baby has with his mother and the dominant modality of objects’ exploration (e.g. through mouthing) during the first months of life [34].
read more
3. What is the role of the robot in the development of its cognitive architecture?
Not only should the robot be able to develop its prospection and action space incrementally and autonomously, but it should be capable of operating in a social environment, profiting of humans to improve its knowledge.
read more
4. What is the role of the robot in learning from humans?
interaction is essentially physical: babies exploit the physical support of their environment, manipulate objects, use physical contact as a means for learning from humans.
read more
![Fig. 10. The active exploration strategy in the human-robot interaction scenario: at each decision step, the exploration algorithm determines the triple (object, action, actor), that is the object to explore, the action to perform on it, and the actor who is going to do the action (either robot or human). The choice of the triple can be done by the human caregiver if the robot is completely passive as in [71], or by the intrinsic motivation system if the robot is active. Top: when the exploration is active, the robot decides to manipulate the object, and can choose for example between two different actions: pushing or throwing the object. A pushing sequence consists in reaching the object on one side then pushing it in a direction parallel to the y-axis of the robot (e.g. towards the left with the right arm). A throwing sequence consists in reaching the object from the top, lifting it then opening the hand at once to release the object. The object drop is assimilated to a throwing, but of course it is more controlled and reasonably keeps the fallen object in the robot’s workspace. As objects fall, their appearance changes quite unpredictably. Bottom: during active exploration, the robot can decide to ask the human to show a new object, or to manipulate the current one. In the first case, the human simply intervenes to change the object on the table. In the second case, the human can act on the object. If the human is a “good teacher”, he can change radically the appearance of the object, for example by flipping it or putting it on one side: this action is beneficial for the robot, because the robot can have a new experience of the object, take more views, etc. If the human is a “bad teacher”, he simply moves the object without caring to change its appearance to the robot.](/figures/fig-10-the-active-exploration-strategy-in-the-human-robot-24ef78sb.webp)
