Know Your Body Through Intrinsic Goals

Abstract

The first “object” that newborn children play with is their own body. This activity allows them to autonomously form a sensorimotor map of their own body and a repertoire of actions supporting future cognitive and motor development. Here we propose the theoretical hypothesis, operationalized as a computational model, that this acquisition of body knowledge is not guided by random motor-babbling, but rather by autonomously generated goals formed on the basis of intrinsic motivations. Motor exploration leads the agent to discover and form representations of the possible sensory events it can cause with its own actions. When the agent realizes the possibility of improving the competence to re-activate those representations, it is intrinsically motivated to select and pursue them as goals. The model is based on four components: (1) a self-organizing neural network, modulated by competence-based intrinsic motivations, that acquires abstract representations of experienced sensory (touch) changes; (2) a selector that selects the goal to pursue, and the motor resources to train to pursue it, on the basis of competence improvement; (3) an echo-state neural network that controls and learns, through goal-accomplishment and competence, the agent's motor skills; (4) a predictor of the accomplishment of the selected goals generating the competence-based intrinsic motivation signals. The model is tested as the controller of a simulated simple planar robot composed of a torso and two kinematic 3-DoF 2D arms. The robot explores its body covered by touch sensors by moving its arms. The results, which might be used to guide future empirical experiments, show how the system converges to goals and motor skills allowing it to touch the different parts of own body and how the morphology of the body affects the formed goals. The convergence is strongly dependent on competence-based intrinsic motivations affecting not only skill learning and the selection of formed goals, but also the formation of the goal representations themselves.

Action-outcome contingencies as the engine of open-ended learning: computational models and developmental experiments

“…For lack of space, here we describe the model at a high level which is however enough to indicate a possible way in which the hypothesis incorporated in the blueprint architecture can be implemented into a specific model. Further details on the system can be found in the companion paper [28].…”

“…II). A second contribution is an illustration of how the blueprint architecture can guide the construction of specific computational models: this is done by overviewing a model illustrated in detail in a companion paper [28] (Sec. III).…”

Prerequisites for an Artificial Self

“…Interesting advances have been made also in the context of goal generation. For instance, Mannella et al (2018) show how an artificial system can autonomously generate goals to be used in an intrinsic motivation system to explore and to gather knowledge about its own body. In Schillaci et al (2020), the authors present an architecture for curiosity-driven goal-directed exploration behaviors on a camera-equipped robot arm.…”

Sensorimotor contingencies as a key drive of development: from babies to robots

“…The way peripersonal motor maps can help robot learning has been studied in Jamone et al (2014). Mannella et al (2018) presented a specific hypothesis on the learning of sensorimotor contingencies in relation to the gradual acquisition of the knowledge of the agent's own body driven by intrinsic motivations. In particular, a simulated robotic model of a torso with two arms is free to move around with its effectors.…”

“…The value of the architecture is that it proposes a possible way to integrate such processes, thus offering a coherent view of how they could work together. The architecture can also inform the construction of specific computational models, as was done in Mannella et al (2018). The architecture is based on the hypothesis that contingency-based learning encompasses general mechanisms able to support the self-generation of goals, the encoding of the perceptual consequences of actions, and the learning of actions of any level of complexity, e.g., from moving a single finger to performing grasping with the whole hand.…”