How to Select a Suitable Action against Strong Pushes in Adult-Size Humanoid Robot: Learning from Past Experiences

2016 ◽  
Author(s):  
S. Mohammadreza Kasaei ◽  
S. Hamidreza Kasaei ◽  
Ehsan Shahri ◽  
Ali Ahmadi ◽  
Nuno Lau ◽  
...  
Keyword(s):  
Humanoid Robot ◽  
Robot Learning ◽  
Adult Size ◽  
Past Experiences

Color Image Processing and Self-Localization for a Humanoid Soccer Robot

2013 ◽  
Vol 461 ◽  
pp. 877-885
Author(s):  
Wei Hong ◽  
Yan Hui Zhang ◽  
Yan Tao Tian ◽  
Chang Jiu Zhou
Keyword(s):  
Image Processing ◽  
Humanoid Robot ◽  
Color Image ◽  
Hopfield Network ◽  
Color Image Segmentation ◽  
Adult Size ◽  
Image Processing Algorithm ◽  
Soccer Robot ◽  
Operator Field ◽  
Rule Changes

The paper proposed a series of image processing algorithm to recognize the evidences in an image accurately for humanoid soccer robot, such as color image segmentation based on HSV model, edge detection based on four linear operator, field straight line extraction by Hough transform based on 8-neighbhour connected domain clusters and identification of line intersection shape based on Hopfield network. Based on evidences from image processing, Piecewise Monte Carlo localization is presented to solve kidnap problem so that localization of humanoid robot can be not only adapt to rule changes for competition, but also be more efficient and robust. The effectiveness of the piecewise MCL is verified by RoboCup Adult Size humanoid soccer robot, Erectus. The experimental results showed that the humanoid robot was able to solve the kidnap problem adaptively with two strategies: resetting or revising, in which resetting was more efficient than revising gradually.

scholarly journals LEARNING TO ACT FROM OBSERVATION AND PRACTICE

2004 ◽  
Vol 01 (04) ◽  
pp. 585-611 ◽  
Author(s):  
DARRIN C. BENTIVEGNA ◽  
CHRISTOPHER G. ATKESON ◽  
ALEŠ UDE ◽  
GORDON CHENG
Keyword(s):  
Domain Knowledge ◽  
Humanoid Robot ◽  
Humanoid Robots ◽  
Robot Learning ◽  
Learning Problems

We present a method for humanoid robots to quickly learn new dynamic tasks from observing others and from practice. Ways in which the robot can adapt to initial and also changing conditions are described. Agents are given domain knowledge in the form of task primitives. A key element of our approach is to break learning problems up into as many simple learning problems as possible. We present a case study of a humanoid robot learning to play air hockey.

DYNAMIC ROLL-AND-RISE MOTION BY AN ADULT-SIZE HUMANOID ROBOT

2004 ◽  
Vol 01 (03) ◽  
pp. 497-516 ◽  
Author(s):  
YASUO KUNIYOSHI ◽  
YOSHIYUKI OHMURA ◽  
KOJI TERADA ◽  
AKIHIKO NAGAKUBO
Keyword(s):  
Critical Points ◽  
Humanoid Robot ◽  
Humanoid Robots ◽  
The Body ◽  
Whole Body ◽  
Adult Size ◽  
Constrained Motion ◽  
Goal State ◽  
Trade Offs ◽  
Alternative Approach

Whole-body dynamic actions under various contacts with the environment will be very important for future humanoid robots to support human tasks in unstructured environments. Such skills are very difficult to realize using the standard motion control methodology based on asymptotic convergence to the successive desired states. An alternative approach would be to exploit the passive dynamics of the body under constrained motion, and to navigate through multiple dynamics by imposing the least control in order to robustly reach the goal state. As a first example of such a strategy, we propose and investigate a "Roll-and-Rise" motion. This is a fully dynamic whole-body task including underactuated motion whose state trajectory is insoluble, and unpredictable perturbations due to complex contacts with the ground. First, we analyze the global structure of Roll-and-Rise motion. Then the critical points are analyzed using simplified models and simulations. The results suggest a non-uniform control strategy which focuses on sparse critical points in the global phase space, and allows deviations and trade-offs at other parts. Finally, experiments with a real adult-size humanoid robot are successfully carried out. The robot rose from a flat-lying posture to a crouching posture within 2 seconds.

scholarly journals Dynamic Roll-and-Rise Motion by an Adult-Size Humanoid Robot

2005 ◽  
Vol 23 (6) ◽  
pp. 706-717 ◽  
Author(s):  
Yasuo Kuniyoshi ◽  
Yoshiyuki Ohmura ◽  
Koji Terada ◽  
Akihiko Nagakubo
Keyword(s):  
Humanoid Robot ◽  
Adult Size

VISION-BASED OBSTACLE AVOIDANCE NAVIGATION WITH AUTONOMOUS HUMANOID ROBOTS FOR STRUCTURED COMPETITION PROBLEMS

2013 ◽  
Vol 10 (03) ◽  
pp. 1350021 ◽  
Author(s):  
CHUNG-HSIEN KUO ◽  
HUNG-CHYUN CHOU ◽  
SHOU-WEI CHI ◽  
YU-DE LIEN
Keyword(s):  
Obstacle Avoidance ◽  
Humanoid Robot ◽  
Ground Truth ◽  
Humanoid Robots ◽  
Visual Localization ◽  
Adult Size ◽  
Map Construction ◽  
Rule Approach ◽  
High Level ◽  
Autonomous Humanoid

Biped humanoid robots have been developed to successfully perform human-like locomotion. Based on the use of well-developed locomotion control systems, humanoid robots are further expected to achieve high-level intelligence, such as vision-based obstacle avoidance navigation. To provide standard obstacle avoidance navigation problems for autonomous humanoid robot researches, the HuroCup League of Federation of International Robot-Soccer Association (FIRA) and the RoboCup Humanoid League defined the conditions and rules in competitions to evaluate the performance. In this paper, the vision-based obstacle avoidance navigation approaches for humanoid robots were proposed in terms of combining the techniques of visual localization, obstacle map construction and artificial potential field (APF)-based reactive navigations. Moreover, a small-size humanoid robot (HuroEvolutionJR) and an adult-size humanoid robot (HuroEvolutionAD) were used to evaluate the performance of the proposed obstacle avoidance navigation approach. The navigation performance was evaluated with the distance of ground truth trajectory collected from a motion capture system. Finally, the experiment results demonstrated the effectiveness of using vision-based localization and obstacle map construction approaches. Moreover, the APF-based navigation approach was capable of achieving smaller trajectory distance when compared to conventional just-avoiding-nearest-obstacle-rule approach.

A Humanoid Robot Learning Audiovisual Classification By Active Exploration

2021 ◽  
Author(s):  
Glareh Mir ◽  
Matthias Kerzel ◽  
Erik Strahl ◽  
Stefan Wermter
Keyword(s):  
Humanoid Robot ◽  
Robot Learning ◽  
Active Exploration

scholarly journals Acquisition and Utilization of Mental Imagery Capability in Robotic Action Sequencing Tasks

2020 ◽  
Vol 13 (2) ◽  
pp. 196-207
Author(s):  
Kristsana Seepanomwan
Keyword(s):  
Mental Imagery ◽  
Tool Use ◽  
Humanoid Robot ◽  
Computational Models ◽  
Robot Learning ◽  
Service Robot ◽  
Human Infants ◽  
Action Sequencing ◽  
Robotic Application ◽  
Robotic Applications

This work presents a series of neurorobotic models underlying learning in robots. It demonstrates the way in which, during sensorimotor exploration, robots do not just gain knowledge about how to form movement primitives but also obtain a mental imagery capability. Mental imagery plays a key role in these computational models by accelerating learning processes of action sequencing tasks. The first experiment involves permitting a humanoid robot to learn how to retrieve an out-of-reach object using a provided tool. This experiment explores a phenomenon on tool use development found in human infants. In addition, it tests two hypotheses on tool use development. The second experiment extends the domain of robot learning by targeting a useful robotic application. It drives a service robot to learn to acquire knowledge of how to manipulate perceived objects based on the objects’ information and a goal from users. By means of planning, learning the sequence of actions in mind, the robots are able to learn by examining actions’ outcome without really performing actions. This allows the second model to completely exclude parts of overt movements from the training loop. The results confirm that two types of robots can complete their given tasks in a reasonable period of time. The proposed models would benefit robotic applications in terms of online learning.

Sign in / Sign up

Export Citation Format

Share Document