405 Hits in 9.7 sec

Value constrained model-free continuous control [article]

Steven Bohez, Abbas Abdolmaleki, Michael Neunert, Jonas Buchli, Nicolas Heess, Raia Hadsell
2019 arXiv   pre-print
We demonstrate the efficacy of our approach using a number of continuous control benchmark tasks, a realistic, energy-optimized quadruped locomotion task, as well as a reaching task on a real robot arm  ...  The naive application of Reinforcement Learning algorithms to continuous control problems -- such as locomotion and manipulation -- often results in policies which rely on high-amplitude, high-frequency  ...  swingup, humanoid stand and humanoid walk.  ... 
arXiv:1902.04623v1 fatcat:7cwedpwtjneg3jotb5xmk2ay5i

NIMBRO-OP2X: Affordable Adult-sized 3D-printed Open-Source Humanoid Robot for Research

Grzegorz Ficht, Hafez Farazi, Diego Rodriguez, Dmytro Pavlichenko, Philipp Allgeuer, Andre Brandenburger, Sven Behnke
2020 International Journal of Humanoid Robotics  
For several years, high development and production costs of humanoid robots restricted researchers interested in working in the field.  ...  In this paper, we present the NimbRo-OP2X, a capable and affordable adult-sized humanoid platform aiming to significantly lower the entry barrier for humanoid robot research.  ...  Acknowledgements This work was partially funded by grant BE 2556/13 of the German Research Foundation (DFG).  ... 
doi:10.1142/s0219843620500218 fatcat:habn2fyoqbg63emvpghdn3uitq

Multipurpose Low-Cost Humanoid Platform and Modular Control Software Development [chapter]

Filipe Silva, Viacutetor Santos
2007 Humanoid Robots, Human-like Machines  
Acknowledgments The authors would like to thank the following students at the University of Aveiro for their support in the humanoid hardware and software development: David Gameiro, Filipe Carvalho, Luis  ...  Naturally, efficient power usage was also a concern and adequate power autonomy had to be sought.  ...  The solution adopted was to use toothed belts to connect gears with a reduction ratio of 2.5:1 ( Figure 3 ).  ... 
doi:10.5772/4798 fatcat:5qwvzcd2jve7hixac6szhamffu

Sound-event recognition with a companion humanoid

Maxime Janvier, Xavier Alameda-Pineda, Laurent Girinz, Radu Horaud
2012 2012 12th IEEE-RAS International Conference on Humanoid Robots (Humanoids 2012)  
The method is validated with a large set of audio data recorded with a humanoid robot in a house.  ...  We present a complete system that takes as input a continuous signal, splits it into significant isolated sounds and noise, and classifies the isolated sounds using a catalogue of learned sound-event classes  ...  Formally, a VQ quantizer is thus a function q that maps a feature vector space U to acodebook C according to: q : U −→ C, q(u) = argmin c∈C d(u, c), (2) where d stands here for the Euclidean distance.  ... 
doi:10.1109/humanoids.2012.6651506 dblp:conf/humanoids/JanvierAGH12 fatcat:mbalqnwf4newndvmeapqrlj3ge

An energy efficient dynamic gait for a Nao robot

Zhenglong Sun, Nico Roos
2014 2014 IEEE International Conference on Autonomous Robot Systems and Competitions (ICARSC)  
We use policy gradient reinforcement learning to identify the optimal parameters of the new gait for a Nao humanoid robot.  ...  This paper presents a framework to generate energy efficient dynamic human-like walk for a Nao humanoid robot.  ...  Fig. 5 shows the energy consumption of the three gaits. We see that the learned gait results in a power reduction of 44% of standard gait and a reduction of 23.8% of hand-tune gait. VII.  ... 
doi:10.1109/icarsc.2014.6849797 dblp:conf/icarsc/SunR14 fatcat:khyxt3wryvh2rm2bwbjssbdr6e

Sample Efficient Reinforcement Learning via Model-Ensemble Exploration and Exploitation [article]

Yao Yao, Li Xiao, Zhicheng An, Wanpeng Zhang, Dijun Luo
2021 arXiv   pre-print
However, there are some remaining issues, such as planning efficient explorations to learn more accurate dynamic models, evaluating the uncertainty of the learned models, and more rational utilization  ...  of models.  ...  By doing so, VIME [38] was proposed to maximize the information gain in the learning process of the prediction model, measured by the reduction of entropy.  ... 
arXiv:2107.01825v1 fatcat:fkzgnwkaxbblxkgn4jgqcfwqni

Measuring Progress in Deep Reinforcement Learning Sample Efficiency [article]

Florian E. Dorner
2021 arXiv   pre-print
Sampled environment transitions are a critical input to deep reinforcement learning (DRL) algorithms.  ...  As simulating real world processes is often prohibitively hard and collecting real world experience is costly, sample efficiency is an important indicator for economically relevant applications of DRL.  ...  Acknowledgments This work was in part completed as a Summer Research Fellow at the Future of Humanity Institute.  ... 
arXiv:2102.04881v1 fatcat:szmlx6yzl5ac3npmknmapghvay


Petar Kormushev, Edwin Olson, Ashutosh Saxena, Wataru Takano
2012 Cybernetics and Information Technologies  
Overview of the mechanical model COMAN (stands for COmpliant huMANoid) is powered by series elastic actuators and is being developed within the AMARSI European project The coupled mechanical model of the  ...  . , q q and q are positions, velocities and accelerations of all joints in vector form. Similarly, , m m q q and m q are the positions, velocities and accelerations of the motors.  ...  However, designing walking gaits by hand for legged robots is a difficult and time-consuming process, so we seek algorithms for learning such gaits to automatically using real world experimentation.  ... 
doi:10.2478/cait-2012-0016 fatcat:ikosrhqlobe7da64pgprx554me

Tracking body motions in order to guide a robot using the time of flight technology

Feryel Zoghlami, Harald Heinrich, Germar Schneider, Mohamed Ali Hamdi
2019 International Conference on Indoor Positioning and Indoor Navigation  
We present in the first part a depth Time of Flight (ToF) based camera solution using machine learning for human body parts recognition and localization in the 3-dimensional space.  ...  In the second part, we use this information to teach robot movements and control its behavior in real time. Our solution recognizes with an accuracy of 90% the hand.  ...  Saxony, the German Federal Ministry of Education and national funding authorities from involved countries.  ... 
dblp:conf/ipin/ZoghlamiHSH19 fatcat:jlrjibwpf5en3cfnye44yd4qpa

Improving Sample Efficiency in Evolutionary RL Using Off-Policy Ranking [article]

Eshwar S R, Shishir Kolathaya, Gugan Thoppe
2022 arXiv   pre-print
This leads to a lot of wasteful interactions since, once the ranking is done, only the data associated with the top-ranked policies is used for subsequent learning.  ...  Evolution Strategy (ES) is a powerful black-box optimization technique based on the idea of natural evolution.  ...  This survival rewards cause the ARS and TRES algorithms to learn policies that stand still for long.  ... 
arXiv:2208.10583v1 fatcat:ndk6a3cn2fh35eoggbj3x2wi5m

Hierarchical spectro-temporal features for robust speech recognition

Xavier Domont, Martin Heckmann, Frank Joublin, Christian Goerick
2008 Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing  
The whitening and dimension reduction are obtained by the following linear transform: x = Q k v, (5.9) where the k × n matrix Q k is given by: Q k = ⎡ ⎢ ⎣ 1 σ1 · · · 0 0 · · · 0 . . . . . . . . . . . .  ...  From an engineering point of view, it means that we can save memory and processing power by using less frequency bins for the high frequencies than for the low frequencies.  ... 
doi:10.1109/icassp.2008.4518635 dblp:conf/icassp/DomontHJG08 fatcat:c7zczp3pvnfizf6knw2hbnsa3m

Learning a Low-dimensional Representation of a Safe Region for Safe Reinforcement Learning on Dynamical Systems [article]

Zhehua Zhou, Ozgur S. Oguz, Marion Leibold, Martin Buss
2020 arXiv   pre-print
Based on the simplified system model, a low-dimensional representation of the safe region is identified and is used to provide safety estimates for learning algorithms.  ...  For safely applying reinforcement learning algorithms on high-dimensional nonlinear dynamical systems, a simplified system model is used to formulate a safe reinforcement learning framework.  ...  Comparison with Physically Inspired Model Order Reduction We compare the proposed approach with the physically inspired model order reduction presented in [18] in terms of representation power of the  ... 
arXiv:2010.09555v1 fatcat:vw2jhvcp4bfsrlozdpowecofme

Detection of Landmines Using Nuclear Quadrupole Resonance (NQR): Signal Processing to Aid Classification [chapter]

S. D. Somasundaram, K. Althoefer, J. A. S. Smith, L. D. Seneviratne
2006 Climbing and Walking Robots  
This laboratory is devoted to autonomy of humanoid robots and uses as testbed the robot HRP-2. The objective of the authors is to achieve a vertical jump in experimentation using this humanoid robot.  ...  Abstract This paper presents the design process and the dynamic analysis of the 21 DOF humanoid robot RH-0, developed in the Robotics Lab. at the University Carlos III of Madrid.  ...  We explain how we use a simulated annealing algorithm for calibrating different parametric sets for different ranges of speeds, with the constriction of avoiding discontinuities, and we present our results  ... 
doi:10.1007/3-540-26415-9_100 dblp:conf/clawar/SomasundaramASS05 fatcat:4vu25zlx65dszgbcxkusxozfve

Exploring the Use of Static and Dynamic Analysis to Improve the Performance of the Mining Sandbox Approach for Android Malware Identification [article]

Francisco Handrick da Costa, Ismael Medeiros, Thales Menezes, João Victor da Silva, Ingrid Lorraine da Silva, Rodrigo Bonifácio, Krishna Narasimhan, Márcio Ribeiro
2021 arXiv   pre-print
Although the use of dynamic analysis for mining Android sandboxes has been empirically investigated before, little is known about the potential benefits of combining static analysis with the mining sandbox  ...  For instance, the first study reveals that a static analysis component of DroidFax (a tool used for instrumenting Android apps in the Bao et al. study) contributes substantially to the performance of the  ...  This might suggest that, since Humanoid learn how humans interact with apps, and use the learned model to guide test generation, at simulate environment, this method to generate test inputs are less  ... 
arXiv:2109.06613v1 fatcat:2ybkuysal5hqhp3kakf54yhqbm

Artificial Intelligence Control System Applied in Smart Grid Integrated Doubly Fed Induction Generator-Based Wind Turbine: A Review

Ramesh Kumar Behara, Akshay Kumar Saha
2022 Energies  
Here, machine learning computations and transfer functions measure physical inertia and control system design's association with power, voltage, and frequency response.  ...  The finding of the review in the paper indicates that artificial intelligence-based machine and deep learning predictive diagnosis fields have gained prominence because of their low cost, less infrastructure  ...  Machine and deep learning methodologies use data to generate intelligent systems that can hold tasks entailing humanoid intellectual [32] .  ... 
doi:10.3390/en15176488 fatcat:vvzk3yypafgh5n74aznvsfrnpe
« Previous Showing results 1 — 15 out of 405 results