96,749 Hits in 4.3 sec

Resonant hopping of a robot controlled by an artificial neural oscillator

Evan H Pelc, Monica A Daley, Daniel P Ferris
2008 Bioinspiration & Biomimetics  
We used a simple hopping robot controlled by an artificial neural oscillator to test the ability of the neural oscillator to adaptively drive this hybrid dynamic system.  ...  The findings provide the first evidence that an artificial neural oscillator will drive a hybrid dynamic system at partial resonance.  ...  This work was supported by an NSF CAREER award to DPF (BES-0347479), an NIH grant to DPF (R01 NS45486), an NSF Bioinformatics Postdoctoral Fellowship to MAD (DBI-0630664) and the University of Michigan  ... 
doi:10.1088/1748-3182/3/2/026001 pmid:18369282 fatcat:qa3zoe275vaejckopw4tn7oc7i

Nonseparable Symplectic Neural Networks [article]

Shiying Xiong, Yunjin Tong, Xingzhe He, Shuqi Yang, Cheng Yang, Bo Zhu
2022 arXiv   pre-print
We demonstrated the efficacy and versatility of our method by predicting a wide range of Hamiltonian systems, both separable and nonseparable, including chaotic vortical flows.  ...  To solve the problem, we propose a novel neural network architecture, Nonseparable Symplectic Neural Networks (NSSNNs), to uncover and embed the symplectic structure of a nonseparable Hamiltonian system  ...  In figure 5 , we use our trained model to predict the dynamics of a 6000-particle systems, including Taylor and Leapfrog vortices.  ... 
arXiv:2010.12636v3 fatcat:y7secy4o6bgalkn7fmlywcavpe

Neuromechanic: A computational platform for simulation and analysis of the neural control of movement

Nathan E. Bunderson, Jeffrey T. Bingham, M. Hongchul Sohn, Lena H. Ting, Thomas J. Burkholder
2012 International Journal for Numerical Methods in Biomedical Engineering  
Explicit inclusion of all state dependencies allows calculation of system derivatives with respect to kinematic states and muscle and neural control states, thus affording a wealth of analytical tools,  ...  The computation of rigid-body dynamics, muscle forces, and activation of the muscles are often performed separately.  ...  This is partly motivated by well-developed formalism for modeling systems of rigid links and by easy access to many excellent rigid body dynamics engines.  ... 
doi:10.1002/cnm.2486 pmid:23027632 pmcid:PMC4347873 fatcat:d572vlhc7vdtpbsbpzl5innf4e

Neural Annealing and Visualization of Autoregressive Neural Networks in the Newman–Moore Model

Estelle M. Inack, Stewart Morawetz, Roger G. Melko
2022 Condensed Matter  
We explore this possibility by implementing the neural annealing method with autoregressive neural networks on a model that exhibits glassy and fractal dynamics: the two-dimensional Newman–Moore model  ...  Artificial neural networks have been widely adopted as ansatzes to study classical and quantum systems.  ...  Note that some regions of low free energy separated by barriers of higher free energy are visible.  ... 
doi:10.3390/condmat7020038 fatcat:wlid6o5fffa6xh6cnzf3bcolvi

Temperature-transferable coarse-graining of ionic liquids with dual graph convolutional neural networks [article]

Jurgis Ruza, Wujie Wang, Daniel Schwalbe-Koda, Simon Axelrod, William H. Harris, Rafael Gomez-Bombarelli
2020 arXiv   pre-print
To improve the coarse-graining of ILs, we present a neural network model trained on all-atom classical molecular dynamics simulations.  ...  Nevertheless, constructing the many-body potential of mean force that defines the structure and dynamics of a coarse-grained system can be complicated and computationally intensive.  ...  Figure 2 shows the neural network architecture for fitting a coarse-grained potential.  ... 
arXiv:2007.14144v1 fatcat:xhidpy5iqzc25iuog6vrjqjwj4

Experimental Studies of Neural Network Control for One-Wheel Mobile Robot

P. K. Kim, S. Jung
2012 Journal of Control Science and Engineering  
To compensate for uncertainties in robot dynamics, a neural network is added to the nonmodel-based PD-controlled system.  ...  Two actuators are used for balancing control by virtue of gyro effect and one actuator for driving movements.  ...  One of the advantages of using a neural network as an auxiliary controller is that the dynamic model of the system is not required.  ... 
doi:10.1155/2012/194397 fatcat:3pdqy5l2zfa7fbhmnj6dvgjdpe

Augmenting Differentiable Simulators with Neural Networks to Close the Sim2Real Gap [article]

Eric Heiden, David Millard, Erwin Coumans, Gaurav S. Sukhatme
2020 arXiv   pre-print
We present a differentiable simulation architecture for articulated rigid-body dynamics that enables the augmentation of analytical models with neural networks at any point of the computation.  ...  applications, while poor local optima are overcome through a random search approach.  ...  In the simulation code, such neural scalars (Figure 2 right) are assigned a unique name, so that in a separate experiment code a "neural blueprint" is defined that declares the neural network architectures  ... 
arXiv:2007.06045v1 fatcat:2jm34beyyvaf7nzeuqayhxd6w4

iLQR-VAE : control-based learning of input-driven dynamics with applications to neural data [article]

Marine Schimel, Ta-Chu Kao, Kristopher T. Jensen, Guillaume Hennequin
2021 bioRxiv   pre-print
To achieve this, a common approach is to record neural populations in behaving animals, and model these data as emanating from a latent dynamical system whose state trajectories can then be related back  ...  Understanding how neural dynamics give rise to behaviour is one of the most fundamental questions in systems neuroscience.  ...  M.S. was funded by an EPSRC DTP studentship and K.T.J. was funded by a Gates Cambridge scholarship.  ... 
doi:10.1101/2021.10.07.463540 fatcat:qwta32ewu5acrenaywi7jdrjxu

Page 214 of Journal of Cognitive Neuroscience Vol. 10, Issue 2 [page]

1998 Journal of Cognitive Neuroscience  
Texture segregation, sur- face representation, and figure-ground separation. Vision Research, in press. Grunewald, A. (1995). Temporal dynamics of visual percep- tion.  ...  Grossberg, S$. (1994). 3-D vision and figure-ground separation by visual cortex. Perception & Psychophysics, 55, 48-120. Grossberg, S. (1995). The attentive brain. American Scientist, 83, 438-449.  ... 

Physics-Incorporated Convolutional Recurrent Neural Networks for Source Identification and Forecasting of Dynamical Systems [article]

Priyabrata Saha, Saurabh Dash, Saibal Mukhopadhyay
2021 arXiv   pre-print
We formulate our model PhICNet as a convolutional recurrent neural network (RNN) which is end-to-end trainable for spatio-temporal evolution prediction of dynamical systems and learns the source behavior  ...  Though the core dynamics follows some principles of physics, real-world physical processes are often driven by unknown external sources.  ...  Saha) Figure 1 : 1 (a): An example dynamical system governed by partially-known PDE and unknown source dynamics.  ... 
arXiv:2004.06243v3 fatcat:ev3u5wf4b5fzbliqpkqojalcga

Time-Reversal Symmetric ODE Network [article]

In Huh, Eunho Yang, Sung Ju Hwang, Jinwoo Shin
2021 arXiv   pre-print
Then, we design a new framework, which we name as Time-Reversal Symmetric ODE Networks (TRS-ODENs), that can learn the dynamics of physical systems more sample-efficiently by learning with the proposed  ...  Time-reversal symmetry, which requires that the dynamics of a system should not change with the reversal of time axis, is a fundamental property that frequently holds in classical and quantum mechanics  ...  represent the ODE functions f in (1) by neural networks and learn the unknown dynamics from data.  ... 
arXiv:2007.11362v3 fatcat:ytfun5zqxresth2h6f2ektoa6u

Dissipative Hamiltonian Neural Networks: Learning Dissipative and Conservative Dynamics Separately [article]

Andrew Sosanya, Sam Greydanus
2022 arXiv   pre-print
We train our model to decompose a damped mass-spring system into its friction and inertial terms and then show that this decomposition can be used to predict dynamics for unseen friction coefficients.  ...  We propose Dissipative Hamiltonian Neural Networks (D-HNNs), which parameterize both a Hamiltonian and a Rayleigh dissipation function.  ...  Figure 1 : 1 Figure 1: Visualizing the architecture of a Dissipative Hamiltonian Neural Network (D-HNN). D-HNNs leverage two neural networks to model dynamic systems.  ... 
arXiv:2201.10085v2 fatcat:e4zhx2w5mncjticka7t2cphz54

Neural Network Based Enveloping Model of Agricultural Tyre

Stojic B.
2019 International Journal of Engineering and Management Sciences  
The behaviour of the tyre in this sense is significantly influenced by the mechanism of geometric filtering of the short-wavelength unevenness of the unprepared ground mostly encountered by tractor during  ...  Based on experimental results, neural network based model of tyre enveloping behaviour was developed.  ...  The optimal choice of the topology of a neural network from the aspect of modelling the regarded system or phenomenon is a problem for whose solution there is no uniquely defined approach.  ... 
doi:10.21791/ijems.2019.1.27. fatcat:gq2hid4wwrbt5bmvmalh6fmlfa

A Timbre-based Approach to Estimate Key Velocity from Polyphonic Piano Recordings

Dasaem Jeong, Taegyun Kwon, Juhan Nam
2018 Zenodo  
To this end, we separate individual notes of polyphonic piano music using non-negative matrix factorization (NMF) and feed them into a neural network that is trained to discriminate the timbre change according  ...  Estimating the key velocity of each note from polyphonic piano music is a highly challenging task. Previous work addressed the problem by estimating note intensity using a polyphonic note model.  ...  ACKNOWLEDGEMENTS This research was supported/partially supported by Samsung Research Funding & Incubation Center for Future Research.  ... 
doi:10.5281/zenodo.1492359 fatcat:rvoqsw6a2vdzfheqgppvmvlmkm

A dynamic neural field model of temporal order judgments

Lauren N. Hecht, John P. Spencer, Shaun P. Vecera
2015 Journal of Experimental Psychology: Human Perception and Performance  
Temporal ordering of events is biased, or influenced, by perceptual organizationfigure-ground organizationand by spatial attention.  ...  The model posits that neural populations processing the figure are more active, resulting in a peak of activation that quickly builds toward a detection threshold when the onset of a target is presented  ...  But how does a neural system detect the offset of information?  ... 
doi:10.1037/xhp0000097 pmid:26280270 fatcat:jnwfdkx3vnal5onnidf5ugcywa
« Previous Showing results 1 — 15 out of 96,749 results