Maximum entropy signal reconstruction with neural networks

1992 ◽  
Vol 3 (2) ◽  
pp. 195-201 ◽  
Author(s):  
D. Ingman ◽  
Y. Merlis
Keyword(s):  
Neural Networks ◽  
Maximum Entropy ◽  
Signal Reconstruction

scholarly journals Local Correlation and Entropy Maps as Tools for Detecting Defects in Industrial Images

2008 ◽  
Vol 18 (1) ◽  
pp. 41-47 ◽  
Author(s):  
Ewa Skubalska-Rafajłowicz
Keyword(s):  
Neural Networks ◽  
Maximum Entropy ◽  
Gray Level ◽  
Rbf Neural Networks ◽  
Nonparametric Estimator ◽  
Local Correlation ◽  
Local Entropy ◽  
Low Contrast ◽  
Speed Up ◽  
Different Parts

Local Correlation and Entropy Maps as Tools for Detecting Defects in Industrial ImagesThe aim of this paper is to propose two methods of detecting defects in industrial products by an analysis of gray level images with low contrast between the defects and their background. An additional difficulty is the high nonuniformity of the background in different parts of the same image. The first method is based on correlating subimages with a nondefective reference subimage and searching for pixels with low correlation. To speed up calculations, correlations are replaced by a map of locally computed inner products. The second approach does not require a reference subimage and is based on estimating local entropies and searching for areas with maximum entropy. A nonparametric estimator of local entropy is also proposed, together with its realization as a bank of RBF neural networks. The performance of both methods is illustrated with an industrial image.

scholarly journals Signal recognition and background suppression by matched filters and neural networks for Tunka-Rex

2019 ◽  
Vol 216 ◽  
pp. 02003
Author(s):  
D. Shipilov ◽  
P.A. Bezyazeekov ◽  
N.M. Budnev ◽  
D. Chernykh ◽  
O. Fedorov ◽  
...  
Keyword(s):  
Neural Networks ◽  
Antenna Arrays ◽  
Signal Reconstruction ◽  
Cosmic Ray ◽  
Matched Filtering ◽  
Background Suppression ◽  
Signal Recognition ◽  
Air Showers ◽  
Matched Filters ◽  
Prospects Of Application

The Tunka Radio Extension (Tunka-Rex) is a digital antenna array, which measures radio emission of the cosmic-ray air-showers in the frequency band of 30-80 MHz. Tunka-Rex is co-located with the TAIGA experiment in Siberia and consists of 63 antennas, 57 of them are in a densely instrumented area of about 1 km2. In the present workwe discuss the improvements of the signal reconstruction applied for Tunka-Rex. At the first stage we implemented matched filtering using averaged signals as template. The simulation study has shown that matched filtering allows one to decrease the threshold of signal detection and increase its purity. However, the maximum performanceof matched filtering is achievable only in case of white noise, while in reality the noise is not fully random due to different reasons. To recognize hidden features of the noise and treat them, we decided to use convolutional neural network with autoencoder architecture. Taking the recorded trace as an input, the autoencoder returns denoised traces, i.e. removes all signal-unrelated amplitudes. We present the comparison between the standard method of signal reconstruction, matched filtering and the autoencoder, and discuss the prospects of application of neural networks for lowering the threshold of digital antenna arrays for cosmic-ray detection.

scholarly journals Neuronal Ensemble Decoding Using a Dynamical Maximum Entropy Model

2014 ◽  
Vol 2014 ◽  
pp. 1-10 ◽  
Author(s):  
Duho Sin ◽  
Jinsoo Kim ◽  
Jee Hyun Choi ◽  
Sung-Phil Kim
Keyword(s):  
Neural Networks ◽  
Maximum Entropy ◽  
Brain Regions ◽  
Time Varying ◽  
Maximum Entropy Model ◽  
Firing Activity ◽  
Entropy Model ◽  
Neuronal Ensemble ◽  
Neural Information ◽  
Ensemble Activity

As advances in neurotechnology allow us to access the ensemble activity of multiple neurons simultaneously, many neurophysiologic studies have investigated how to decode neuronal ensemble activity. Neuronal ensemble activity from different brain regions exhibits a variety of characteristics, requiring substantially different decoding approaches. Among various models, a maximum entropy decoder is known to exploit not only individual firing activity but also interactions between neurons, extracting information more accurately for the cases with persistent neuronal activity and/or low-frequency firing activity. However, it does not consider temporal changes in neuronal states and therefore would be susceptible to poor performance for nonstationary neuronal information processing. To address this issue, we develop a novel decoder that extends a maximum entropy decoder to take time-varying neural information into account. This decoder blends a dynamical system model of neural networks into the maximum entropy model to better suit for nonstationary circumstances. From two simulation studies, we demonstrate that the proposed dynamic maximum entropy decoder could cope well with time-varying information, which the conventional maximum entropy decoder could not achieve. The results suggest that the proposed decoder may be able to infer neural information more effectively as it exploits dynamical properties of underlying neural networks.

scholarly journals Defect Quantification with Thermographic Signal Reconstruction and Artificial Neural Networks

2006 ◽  
Author(s):  
Hernan Benitez ◽  
Clemente Ibarra-Castanedo ◽  
Humberto Loaiza ◽  
Eduardo Caicedo ◽  
Abdelhakim Bendada ◽  
...  
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Signal Reconstruction ◽  
Artificial Neural ◽  
Defect Quantification
Sign in / Sign up

Export Citation Format

Share Document