Emulation and behavior understanding through shared values

Takahashi, Yasutake; Kawamata, Teruyasu; Asada, Minoru; Negrello, Mario

doi:10.1109/iros.2007.4399385

Cited by 13 publications

(5 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Sometimes, it is the observer who will manually set the reward value 2 ; it can be de¯ned as a mathematical function that maps from states and actions to rewards 17,38,39,29 or the reward function can be learned from the demonstration set, what is called inverse RL. 40 We used stability as a criterion because this task involves moving from a statically stable posture (seated) to an unstable¯xed point (upright posture) through a process of dynamic stabilizing.…”

Section: Generalization and Discussion Of The Reward Profilementioning

confidence: 99%

See 1 more Smart Citation

Full-Body Postural Control of a Humanoid Robot with Both Imitation Learning and Skill Innovation

González-Fierro

Balaguer

Swann

et al. 2014

Int. J. Human. Robot.

View full text Add to dashboard Cite

In this paper, we present a novel methodology to obtain imitative and innovative postural movements in a humanoid based on human demonstrations in a di®erent kinematic scale. We collected motion data from a group of human participants standing up from a chair. Modeling the human as an actuated 3-link kinematic chain, and by de¯ning a multiobjective reward function of zero moment point and joint torques to represent the stability and e®ort, we computed reward pro¯les for each demonstration. Since individual reward pro¯les show vari-ability across demonstrating trials, the underlying state transition probabilities were modeled using a Markov chain. Based on the argument that the reward pro¯les of the robot should show the same temporal structure of those of the human, we used di®erential evolution to compute a trajectory that ¯ts all humanoid constraints and minimizes the di®erence between the robot reward pro¯le and the predicted pro¯le if the robot imitates the human. Therefore, robotic imitation involves developing a policy that results in a temporal reward structure, matching that of a group of human demonstrators across an array of demonstrations. Skill innovation was achieved by optimizing a signed reward error after imitation was achieved. Experimental results using the humanoid HOAP-3 are shown.

show abstract

Section: Generalization and Discussion Of The Reward Profilementioning

confidence: 99%

“…Takahashi presents a multi-agent behavior imitation procedure based on RL. 38 Their method can be divided into two phases. First, they recognize an observed behavior through the estimation of the state and action reward, encoding it as a state value function.…”

Section: Comparison With Related Workmentioning

confidence: 99%

Full-Body Postural Control of a Humanoid Robot with Both Imitation Learning and Skill Innovation

González-Fierro

Balaguer

Swann

et al. 2014

Int. J. Human. Robot.

View full text Add to dashboard Cite

show abstract

“…V (s) andV (s) are the state value updated by oneself and the state value estimated through observation, respectively. Takahashi et al [4] showed the Fig. 1.…”

Section: Introductionmentioning

confidence: 94%

“…Takahashi et al [4] proposed a method of not only to learn and execute a variety of behavior but also to recognize behavior of others supposing that the observer has already acquired the values of all kinds of behavior the observed agent can do. The recognition means, in this paper, that the robot categorizes the observed behavior to a set of its own behavior acquired beforehand.…”

Section: Introductionmentioning

confidence: 99%

Mutual Development of Behavior Acquisition and Recognition based on Value System

Takahashi¹,

Tamura²,

Asada³

2009

J. SOFT

Self Cite

View full text Add to dashboard Cite

Abstract-Both self-learning architecture (embedded structure) and explicit/implicit teaching from other agents (environmental design issue) are necessary not only for one behavior learning but more seriously for life-time behavior learning. This paper presents a method for a robot to understand unfamiliar behavior shown by others through the collaboration between behavior acquisition and recognition of observed behavior, where the state value has an important role not simply for behavior acquisition (reinforcement learning) but also for behavior recognition (observation). That is, the state value updates can be accelerated by observation without real trial and error while the learned values enrich the recognition system since it is based on estimation of the state value of the observed behavior. The validity of the proposed method is shown by applying it to a dynamic environment where two robots play soccer.

show abstract

“…Suppose that the observer has already acquired the utilities (state values in a reinforcement learning scheme). Takahashi et al [52] show that the observer can understand/recognize behaviors shown by a demonstrator based not on a precise object tra- Fig. 10 Affetto: facial expressions, internal structure of the upper torso, and recent pictures jectory in allocentric/egocentric coordinate space but rather on an estimated utility transition during the observed behavior.…”

Section: Perspective-taking Theory Of Mind and Emotion Controlmentioning

confidence: 99%

Towards Artificial Empathy

Asada

2014

Int J of Soc Robotics

Self Cite

View full text Add to dashboard Cite

The design of artificial empathy is one of the most essential issues in social robotics. This is because empathic interactions with ordinary people are needed to introduce robots into our society. Several attempts have been made for specific situations. However, such attempts have provided several limitations; thus, diminishing authenticity. The present article proposes "affective developmental robotics (hereafter, ADR)," which provides more authentic artificial empathy based on the concept of cognitive developmental robotics (hereafter, CDR). First, the evolution and development of empathy as revealed in neuroscience and biobehavioral studies are reviewed, moving from emotional contagion to envy and schadenfreude. These terms are then reconsidered from the ADR/CDR viewpoint, particularly along the developmental trajectory of self-other cognition. Next, a conceptual model of artificial empathy is proposed based on an ADR/CDR viewpoint and discussed with respect to several existing studies. Finally, a general discussion and proposals for addressing future issues are given.

show abstract

Emulation and behavior understanding through shared values

Cited by 13 publications

References 13 publications

Full-Body Postural Control of a Humanoid Robot with Both Imitation Learning and Skill Innovation

Full-Body Postural Control of a Humanoid Robot with Both Imitation Learning and Skill Innovation

Mutual Development of Behavior Acquisition and Recognition based on Value System

Towards Artificial Empathy

Contact Info

Product

Resources

About