Open Access. Powered by Scholars. Published by Universities.®
- Discipline
- Institution
- Publication
- Publication Type
Articles 1 - 12 of 12
Full-Text Articles in Signal Processing
Implementation Of Large Neural Networks Using Decomposition, Henry Selvaraj, H. Niewiadomski, P. Buciak, M. Pleban, Piotr Sapiecha, Tadeusz Luba, Venkatesan Muthukumar
Implementation Of Large Neural Networks Using Decomposition, Henry Selvaraj, H. Niewiadomski, P. Buciak, M. Pleban, Piotr Sapiecha, Tadeusz Luba, Venkatesan Muthukumar
Electrical & Computer Engineering Faculty Research
The article presents methods of dealing with huge data in the domain of neural networks. The decomposition of neural networks is introduced and its efficiency is proved by the authors’ experiments. The examinations of the effectiveness of argument reduction in the above filed, are presented. Authors indicate, that decomposition is capable of reducing the size and the complexity of the learned data, and thus it makes the learning process faster or, while dealing with large data, possible. According to the authors experiments, in some cases, argument reduction, makes the learning process harder.
Voice Command Controller, Hoang Nghia Nguyen
Voice Command Controller, Hoang Nghia Nguyen
Theses : Honours
Signal processing technology has been strongly developed and it has attracted interest from scientists and engineers around the world from the last decade. Speech synthesis and speech recognition are particular topic in the field that have been widely used and developed in many different area such as business, controlling, education and entertainment. The project's main objective is to study and develop an application program with the Speech SDK through design and implementation of Tele-Control system based on the commercial product of National Semiconductor: Carrier-Current Transceiver (LM 1893) and Speech development kit (Speech SDK4.0) from Microsoft Corporation. The project is suitable …
An Investigation Of Preliminary Feature Screening Using Signal-To-Noise Ratios, David B. Sumrell
An Investigation Of Preliminary Feature Screening Using Signal-To-Noise Ratios, David B. Sumrell
Theses and Dissertations
A new saliency metric and a new saliency screening method are developed. This new metric, the SN saliency metric, is based upon signal-to-noise ratios, where the signal is provided by a sum of squared weights associated with a given feature, and the noise is based upon a sum of squared weights associated with a reference noise feature which is injected into the data. The resultant metric allows for a direct comparison of the feature of interest with a reference noise feature which is known to be nonsalient. The SN saliency screening method, which uses the SN saliency metric, offers the …
Adaptive And Fixed Wavelet Features For Narrowband Signal Classification, Anthony J. Pohl
Adaptive And Fixed Wavelet Features For Narrowband Signal Classification, Anthony J. Pohl
Theses and Dissertations
The application of the multiresolution analysis developed by Mallat to signal classification by Pati and Krishnaprasad and Szu, et al, is further explored in this thesis. Several different wavelet based feature extraction and classification systems are developed and implemented. Methods which rely on the traditional dyadic wavelet decomposition and on the adaptive wavelet representation are presented. Each of the classification systems is implemented for a labeled data set of narrowband signals. Finally, classification results on the full data set and on low frequency Fourier coefficients are provided as baseline comparisons for our work.
Space Object Identification Using Spatio-Temporal Pattern Recognition, Gary W. Brandstrom
Space Object Identification Using Spatio-Temporal Pattern Recognition, Gary W. Brandstrom
Theses and Dissertations
This thesis is part of a research effort to automate the task of characterizing space objects or satellites based on a sequence of images. The goal is to detect space object anomalies. Two algorithms are considered - the feature space trajectory neural network (FST NN) and hidden Markov model (HMM) classifier. The FST NN was first presented by Leonard Neiberg and David P. Casasent in 1994 as a target identification tool. Kenneth H. Fielding and Dennis W. Ruck recently applied the hidden Markov model classifier to a 3D moving light display identification problem and a target recognition problem, using time …
Time-Frequency Shift-Tolerance And Counterpropagation Network With Applications To Phoneme Recognition, Li-Minn Ang
Time-Frequency Shift-Tolerance And Counterpropagation Network With Applications To Phoneme Recognition, Li-Minn Ang
Theses : Honours
Human speech signals are inherently multi-component non-stationary signals. Recognition schemes for classification of non-stationary signals generally require some kind of temporal alignment to be performed. Examples of techniques used for temporal alignment include hidden Markov models and dynamic time warping. Attempts to incorporate temporal alignment into artificial neural networks have resulted in the construction of time-delay neural networks. The nonstationary nature of speech requires a signal representation that is dependent on time. Time-frequency signal analysis is an extension of conventional time-domain and frequency-domain analysis methods. Researchers have reported on the effectiveness of time-frequency representations to reveal the time-varying nature of …
Color Image Segmentation, Kimberley A. Mccrae
Color Image Segmentation, Kimberley A. Mccrae
Theses and Dissertations
The most difficult stage of automated target recognition ATR is segmentation. Current AFIT segmentation problems include faces and tactical targets previous efforts to segment these objects have used intensity and motion cues. This thesis develops a color preprocessing scheme to be used with the other segmentation techniques. A neural network is trained to identify the color of a desired object, eliminating all but that color from the scene. Gabor correlations and 2D wavelet transformations will be performed on stationary images and 3D wavelet transforms on multispectral data will incorporate color and motion detection into the machine visual system. The thesis …
Etann Hardware Implementation For Radar Emitter Identification, James B. Calvin Jr.
Etann Hardware Implementation For Radar Emitter Identification, James B. Calvin Jr.
Theses and Dissertations
This study investigated classification of 30 radar emitters with 16 signal features using Intel's 80170NX chip, the Electronically Trainable Analog Neural Network (ETANN). Software tools were developed to characterize the ETANN sigmoidal transfer function for use in a custom simulator, known as Neural Graphics. Neural Graphics operates on a Silicon Graphics workstation. The Intel Neural Network Training System simulators were used in early experiments, but were found to be inefficient in training on data used in this research. Using a modified Neural Graphics simulator, single chip and multi-chip experiments were performed to provide benchmark results prior to performing chip-in-loop training. …
Face Recognition With Neural Networks, Dennis L. Krepp
Face Recognition With Neural Networks, Dennis L. Krepp
Theses and Dissertations
This study investigated neural networks for face verification and classification. The research concentrated on developing a neural network based feature extractor and/or classifier to perform authorized user verification in a realistic work environment. Recognition accuracy, system assumptions, training time, and execution time were analyzed to determine the feasibility of a neural network approach. Data was collected using a camcorder and two segmentation schemes: manual segmentation and motion-based, automatic segmentation. Data consisted of over 2000. 32x32 pixel, 8 bit gray scale images of 52 subjects; each subject had two to ten days worth of images collected. Several training and test sets …
Signal Approximation With A Wavelet Neural Network, Charles M. Westphal
Signal Approximation With A Wavelet Neural Network, Charles M. Westphal
Theses and Dissertations
This study investigated the use of Wavelet Neural Networks (WNN) for signal approximation. The particular wavelet function used in this analysis consisted of a summation of sigmoidal functions (a sigmoidal wavelet). The sigmoidal wavelet has the advantage of being easily implemented in hardware via specialized electronic devices like the Intel Electronically Trainable Analog Neural Network (ETANN) chip. The WNN representation allows the determination of the number of hidden-layer nodes required to achieve a desired level of approximation accuracy. Results show that a bandlimited signal can be accurately approximated with a WNN trained with irregularly sampled data. Signal approximation, Wavelet neural …
Infrared Target Recognition, Brian D. Singstock
Infrared Target Recognition, Brian D. Singstock
Theses and Dissertations
In this thesis, three approaches were used for Automatic Target Recognition (ATR). These approaches were shape, moment and Fourier generated features, Karhunen-Loeve transform (KLT) generated features and Discrete Cosine Transform (DCT) generated features. The KLT approach was modelled after the face recognition research by Suarez, AFIT, and Turk and Pentland, MIT. A KLT is taken of a reduced covariance matrix, composed all three classes of targets, and the resulting eigenimages are used to reconstruct the original images. The reconstruction coefficients for each original image are found by taking the dot product of the original image with each eigenimage. These reconstruction …
Feature Extraction For Pose Estimation. A Comparison Between Synthetic And Real Ir Imagery, Donald J. Willis
Feature Extraction For Pose Estimation. A Comparison Between Synthetic And Real Ir Imagery, Donald J. Willis
Theses and Dissertations
This research addressed the problem of pose estimation of three- dimensional objects given their two-dimensional IR imagery and corresponding synthetic (computer-generated) IR imagery. Features and techniques were investigated to find those which may be extendable from computer models to real- world IR imagery. GTSIG and SCNGEN were used to create the synthetic imagery. Silhouette and outline shape moments were explored as optimum features for the comparison. Employing back-propagation with momentum as the training paradigm, a two-hidden-layer neural network was able to determine the base-plane orientation of the synthetic imagery to within 7.5 degrees with better than 90% accuracy. (No conclusive …