Learning of Motor Control from Motor Babbling**This research is supported by CREST, JST.

Aoki, Tatsuya; Nagaoka, Tomoaki; Nagai, Takayuki

doi:10.1016/j.ifacol.2016.10.478

Cited by 9 publications

(5 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By relying on such adaptive maps, humans can learn to perform many motion tasks in an efficient way. These advanced capabilities have motivated many research studies that attempt to artificially replicate such skills in robots and machines [3,4]. Our aim in this work is to develop a bioinspired adaptive computational method to guide the motion of robots with real-time sensory information and a limited amount of data.…”

Section: Introductionmentioning

confidence: 99%

Differential mapping spiking neural network for sensor-based robot control

2021

View full text Add to dashboard Cite

In this work, a spiking neural network (SNN) is proposed for approximating differential sensorimotor maps of robotic systems. The computed model is used as a local Jacobian-like projection that relates changes in sensor space to changes in motor space. The SNN consists of an input (sensory) layer and an output (motor) layer connected through plastic synapses, with inter-inhibitory connections at the output layer. Spiking neurons are modeled as Izhikevich neurons with a synaptic learning rule based on spike timing-dependent plasticity. Feedback data from proprioceptive and exteroceptive sensors are encoded and fed into the input layer through a motor babbling process. A guideline for tuning the network parameters is proposed and applied along with the particle swarm optimization technique. Our proposed control architecture takes advantage of biologically plausible tools of an SNN to achieve the target reaching task while minimizing deviations from the desired path, and consequently minimizing the execution time. Thanks to the chosen architecture and optimization of the parameters, the number of neurons and the amount of data required for training are considerably low. The SNN is capable of handling noisy sensor readings to guide the robot movements in real-time. Experimental results are presented to validate the control methodology with a vision-guided robot.

show abstract

Section: Introductionmentioning

confidence: 99%

Differential mapping spiking neural network for sensor-based robot control

2021

View full text Add to dashboard Cite

show abstract

“…Crying is a highly precise action policy: the infant learns that its own crying causes the parents' various attempts to minimise its uncertainty along a wide range of responses (Pezzulo et al, 2023; see also: e.g., Frierson et al, 2021). The role of undertaking autonomous activities -even random as infants' 'motor babbling' (i.e., the flailing of the limbs; Aoki et al, 2016;Adolph & Hoch, 2019; see also: e.g., Smith & Gasser, 2005;Adolph & Robinson, 2015;Thelen, 1995) -for the development of skilful capabilities (i.e., goal-directed actions), has been shown by work in embodied robotics adopting the AIF. Research on social understanding has instead show the importance of engaging with the material world.…”

Section: Active Inference Goes To Schoolmentioning

confidence: 99%

Active Inference Goes to School. The Importance of Active Learning in the Age of Large Language Models

Di Paolo,

White,

GUÉNIN--CARLUT

et al. 2024

Preprint

View full text Add to dashboard Cite

Human learning essentially involves embodied interactions with the material world. But our worlds now include increasing numbers of powerful and (apparently) disembodied generative AIs. In what follows we ask how best to understand these new (somewhat “alien”, because of their disembodied nature) resources and how to incorporate them in our educational practices. We focus on methodologies that encourage exploration and embodied interactions with ‘prepared’ material environments, such as the carefully organised settings of Montessori education. Using the Active Inference Framework, we approach our questions by thinking about human learning as epistemic foraging and prediction error minimization. We end by arguing that generative AIs should figure naturally as new elements in prepared learning environments by facilitating sequences of precise prediction error enabling trajectories of self-correction. In these ways we anticipate new synergies between (apparently) disembodied and (essentially) embodied forms of intelligence.

show abstract

“…The internal models are trained through two stages. In the first stage, the models are pre-trained using a coarse forward kinematics model where the robot realizes selfexploration through motor babbling [49]. In this stage, all work is done in simulation and no actual movement is conducted.…”

Section: Establishment Of Internal Modelsmentioning

confidence: 99%

Robot Arm Reaching Based on Inner Rehearsal

Wang,

Zou,

Wei

et al. 2023

Biomimetics

View full text Add to dashboard Cite

Robot arm motion control is a fundamental aspect of robot capabilities, with arm reaching ability serving as the foundation for complex arm manipulation tasks. However, traditional inverse kinematics-based methods for robot arm reaching struggle to cope with the increasing complexity and diversity of robot environments, as they heavily rely on the accuracy of physical models. In this paper, we introduce an innovative approach to robot arm motion control, inspired by the cognitive mechanism of inner rehearsal observed in humans. The core concept revolves around the robot’s ability to predict or evaluate the outcomes of motion commands before execution. This approach enhances the learning efficiency of models and reduces the mechanical wear on robots caused by excessive physical executions. We conduct experiments using the Baxter robot in simulation and the humanoid robot PKU-HR6.0 II in a real environment to demonstrate the effectiveness and efficiency of our proposed approach for robot arm reaching across different platforms. The internal models converge quickly and the average error distance between the target and the end-effector on the two platforms is reduced by 80% and 38%, respectively.

show abstract

Learning of Motor Control from Motor Babbling**This research is supported by CREST, JST.

Cited by 9 publications

References 4 publications

Differential mapping spiking neural network for sensor-based robot control

Differential mapping spiking neural network for sensor-based robot control

Active Inference Goes to School. The Importance of Active Learning in the Age of Large Language Models

Robot Arm Reaching Based on Inner Rehearsal

Contact Info

Product

Resources

About