In this paper, we propose a design of a robotic ankle-foot complex based on the human functional-anatomic ankle-foot structure. The proposed foot consists of three links, two joints, and four plantar muscles, whose mechanical stiffness can be controlled by utilizing McKibben pneumatic actuators.With this structure, a deformable medial longitudinal arch in a human foot can be emulated. We developed a musculoskeletal biped robot to which the proposed feet are implemented and measured its walking motion, especially the deformation of the robot foot. It is found that the foot generates a truss mechanism and a windlass mechanism, which are important functions of a human foot for shock absorption and energy storage and reuse. We also conducted a walking experiment with various parameters of a plantar muscle's tonus to see how the tonus affects to ground reaction forces (GRFs) and its walking behavior. It is found that the GRF had two peaks as well as human walking and the shape of the GRF curve changes according to the tonus of the plantar muscle. We analyzed the impulse of GRF, finding out that a truss mechanism and a windlass mechanism works effectively with appropriate tonus of the plantar aponeurosis.

This paper addresses the generation of referring expressions that not only refer to objects correctly but also let humans find them quickly. As a target becomes relatively less salient, identifying referred objects itself becomes more difficult. However, the existing studies regarded all sentences that refer to objects correctly as equally good, ignoring whether they are easily understood by humans. If the target is not salient, humans utilize relationships with the salient contexts around it to help listeners to comprehend it better. To derive this information from human annotations, our model is designed to extract information from the target and from the environment. Moreover, we regard that sentences that are easily understood are those that are comprehended correctly and quickly by humans. We optimized this by using the time required to locate the referred objects by humans and their accuracies. To evaluate our system, we created a new referring expression dataset whose images were acquired from Grand Theft Auto V (GTA V), limiting targets to persons. Experimental results show the effectiveness of our approach. Our code and dataset are available at https://github.com/mikittt/easy-to-understand-REG.

Feline locomotion combines great acrobatic proficiency, unparalleled balance and higher accelerations than other animals. Capable of accelerating from 0 to 100 km h −1 in three seconds, the cheetah (Acinonyx jubatus) is still a mystery which intrigues scientists. Aiming for a better understanding of the source of such higher speeds, we develop a biomimetic platform, where musculoskeletal parameters (range of motion and moment arms) from the biological system can be evaluated with air muscles within a lightweight robotic structure. We performed experiments validating the muscular structure during a treadmill walk, successfully reproducing animal locomotion while adopting an EMG based control method.

The plasticity of the human nervous system allows us to acquire an open-ended repository of sensorimotor skills in adulthood, such as the mastery of tools, musical instruments or sports. How novel sensorimotor skills are learned from scratch is yet largely unknown. In particular, the so-called inverse mapping from goal states to motor states is underdetermined because a goal can often be achieved by many different movements (motor redundancy). How humans learn to resolve motor redundancy and by which principles they explore high-dimensional motor spaces has hardly been investigated. To study this question, we trained human participants in an unfamiliar and redundant visually-guided manual control task. We qualitatively compare the experimental results with simulation results from a population of artificial agents that learned the same task by Goal Babbling, which is an inverse-model learning approach for robotics. In Goal Babbling, goal-related feedback guides motor exploration and thereby enables robots to learn an inverse model directly from scratch, without having to learn a forward model first. In the human experiment, we tested whether different initial conditions (starting positions of the hand) influence the acquisition of motor synergies, which we identified by Principal Component Analysis in the motor space. The results show that the human participants’ solutions are spatially biased towards the different starting positions in motor space and are marked by a gradual co-learning of synergies and task success, similar to the dynamics of motor learning by Goal Babbling. However, there are also differences between human learning and the Goal Babbling simulations, as humans tend to predominantly use Degrees of Freedom that do not have a large effect on the hand position, whereas in Goal Babbling, Degrees of Freedom with a large effect on hand position are used predominantly. We conclude that humans use goal-related feedback to constrain motor exploration and resolve motor redundancy when learning a new sensorimotor mapping, but in a manner that differs from the current implementation of Goal Babbling due to different constraints on motor exploration.

This paper aims to clarify the mechanism of an infant's locomotive development from the viewpoint of cognitive developmental robotics. We built up an infant-sized musculoskeletal robot driven by McKibben pneumatic actuators, which enable the robot to interact with its environment without any problems of mechanical damage and excessive heat in long-term experiments. We applied a learning algorithm based on central pattern generators and optimization method as a way for the robot to acquire its crawling motion. As a result of a developmental experiment, an efficient forward motion was acquired with the proposed method. We discuss how its musculoskeltal body, spinal structure and tonus of the artificial muscles affect its development of crawling behavior.