Signal Processing Commons^™

Speaker Diarization And Identification From Single-Channel Classroom Audio Recording Using Virtual Microphones, Antonio Gomez

Electrical and Computer Engineering ETDs

Speaker identification in noisy audio recordings, specifically those from collaborative learning environments, can be extremely challenging. There is a need to identify individual students talking in small groups from other students talking at the same time. To solve the problem, we assume the use of a single microphone per student group without any access to previous large datasets for training.

This dissertation proposes a method of speaker identification using cross-correlation patterns associated to an array of virtual microphones, centered around the physical microphone. The virtual microphones are simulated by using approximate speaker geometry observed from a video recording. The patterns …

Design & Analysis Of Mixed-Mode Integrated Circuit For Pulse-Shape Discrimination, Bryan Orabutt

McKelvey School of Engineering Graduate Student Theses & Dissertations

In nuclear science experiments it is usually necessary to determine the type of radiation, its energy and direction with considerable accuracy. The detection of neutrons and discriminating them from gamma rays is particularly difficult. A popular method of doing so is to measure characteristics intrinsic to the pulse shape of each radiation type in order to perform pulse-shape discrimination (PSD).

Historically, PSD capable systems have been designed with two approaches in mind: specialized analog circuitry, or digital signal processing (DSP). In this work we propose a PSD capable circuit topology using techniques from both the analog and DSP domains. We …

Machine Learning Classification Of Digitally Modulated Signals, James A. Latshaw

Electrical & Computer Engineering Theses & Dissertations

Automatic classification of digitally modulated signals is a challenging problem that has traditionally been approached using signal processing tools such as log-likelihood algorithms for signal classification or cyclostationary signal analysis. These approaches are computationally intensive and cumbersome in general, and in recent years alternative approaches that use machine learning have been presented in the literature for automatic classification of digitally modulated signals. This thesis studies deep learning approaches for classifying digitally modulated signals that use deep artificial neural networks in conjunction with the canonical representation of digitally modulated signals in terms of in-phase and quadrature components. Specifically, capsule networks are …

State Estimation—Beyond Gaussian Filtering, Haozhan Meng

LSU New Orleans Theses and Dissertations

This dissertation considers the state estimation problems with symmetric Gaussian/asymmetric skew-Gaussian assumption under linear/nonlinear systems. It consists of three parts. The first part proposes a new recursive finite-dimensional exact density filter based on the linear skew-Gaussian system. The second part adopts a skew-symmetric representation (SSR) of distribution for nonlinear skew-Gaussian estimation. The third part gives an optimized Gauss-Hermite quadrature (GHQ) rule for numerical integration with respect to Gaussian integrals and applies it to nonlinear Gaussian filters.

We first develop a linear system model driven by skew-Gaussian processes and present the exact filter for the posterior density with fixed dimensional recursive …

Poly Camtracker, Nickolas W. Ogilvie, Shawn W. Thai

Electrical Engineering

With the onset of the post-pandemic virtual classroom world, more and more classes are attempting hybrid-style instruction. To accomplish that, recordings of lectures are widely used, with presentation-sharing & audio recording capability. However, video recordings of lectures, with the ability to see the board, lab equipment, and the professor, are the optimal way to give the virtual students the best experience. Right now, technology is readily available for static camera recordings of lecture halls, but this may not suffice when it comes to larger lecture halls and lab demonstrations, where directed shots of equipment or the whiteboard are needed in …

Locating Unknown Interference Sources With Time Difference Of Arrival Estimates, Chia Ying Kuo

McKelvey School of Engineering Graduate Student Theses & Dissertations

Adaptive spectrum sharing between different systems and operators is being deployed in order to make use of the wireless spectrum more efficiently. However, when the spectrum is shared, it can create situations in which an operator is unable to determine the identity of an interferer transmitting an unknown signal. This is the situation in which the POWDER testbed found itself in, starting in late 2021. This thesis provides general-purpose tools for operators to locate an unknown signal source in real-world outdoor environments. We used cross-correlation between the signals measured at multiple time-synchronized base stations to estimate the time difference of …

Multicarrier Modulation Using Discrete Fractional Fourier Transform, Amir Raeisi Nafchi

Electrical and Computer Engineering ETDs

The focus of the research was to investigate the application of the discrete fractional Fourier transform (DFRFT) in communication systems. We investigated the compactness of the Gauss-Hermite like eigenvectors of the DFRFT and showed how a multi-carrier modulation system could benefit from it. This led to identifying an affine DFRFT. We proved the circular convolution property for the proposed DFRFT. Using this affine transform, we were able to design an orthogonal frequency division multiplexer (OFDM) communication system. In the process of implementing the OFDM, we developed a method for fast computation of the DFRFT using the chirp-z transform. Using the …

Composite Style Pixel And Point Convolution-Based Deep Fusion Neural Network Architecture For The Semantic Segmentation Of Hyperspectral And Lidar Data, Kevin T. Decker, Brett J. Borghetti

Faculty Publications

Multimodal hyperspectral and lidar data sets provide complementary spectral and structural data. Joint processing and exploitation to produce semantically labeled pixel maps through semantic segmentation has proven useful for a variety of decision tasks. In this work, we identify two areas of improvement over previous approaches and present a proof of concept network implementing these improvements. First, rather than using a late fusion style architecture as in prior work, our approach implements a composite style fusion architecture to allow for the simultaneous generation of multimodal features and the learning of fused features during encoding. Second, our approach processes the higher …

Dc Microgrid Fault Detection Using Multiresolution Analysis Of Traveling Waves, Rudy Montoya

Mechanical Engineering ETDs

Fast detection and isolation of faults in a DC microgrid is of particular importance. Fast tripping protection (i) increases the lifetime of power electronics (PE) switches by avoiding high fault current magnitudes and (ii) enhances the controllability of PE converters. This thesis proposes a traveling wave (TW) based scheme for fast tripping protection of DC microgrids. The proposed scheme utilizes a discrete wavelet transform (DWT) to calculate the high-frequency components of DC fault currents. Multiresolution analysis (MRA) using DWT is utilized to detect TW components for different frequency ranges. The Parseval energy calculated from the MRA coefficients are then used …

Machine Learning Based Medical Image Deepfake Detection: A Comparative Study, Siddharth Solaiyappan, Yuxin Wen

Engineering Faculty Articles and Research

Deep generative networks in recent years have reinforced the need for caution while consuming various modalities of digital information. One avenue of deepfake creation is aligned with injection and removal of tumors from medical scans. Failure to detect medical deepfakes can lead to large setbacks on hospital resources or even loss of life. This paper attempts to address the detection of such attacks with a structured case study. Specifically, we evaluate eight different machine learning algorithms, which include three conventional machine learning methods (Support Vector Machine, Random Forest, Decision Tree) and five deep learning models (DenseNet121, DenseNet201, ResNet50, ResNet101, VGG19) …

Intra-Hour Solar Forecasting Using Cloud Dynamics Features Extracted From Ground-Based Infrared Sky Images, Guillermo Terrén-Serrano

Electrical and Computer Engineering ETDs

Due to the increasing use of photovoltaic systems, power grids are vulnerable to the projection of shadows from moving clouds. An intra-hour solar forecast provides power grids with the capability of automatically controlling the dispatch of energy, reducing the additional cost for a guaranteed, reliable supply of energy (i.e., energy storage). This dissertation introduces a novel sky imager consisting of a long-wave radiometric infrared camera and a visible light camera with a fisheye lens. The imager is mounted on a solar tracker to maintain the Sun in the center of the images throughout the day, reducing the scattering effect produced …

Three Wave Mixing In Epsilon-Near-Zero Plasmonic Waveguides For Signal Regeneration, Nicholas Mirchandani, Mark C. Harrison

Engineering Faculty Articles and Research

Vast improvements in communications technology are possible if the conversion of digital information from optical to electric and back can be removed. Plasmonic devices offer one solution due to optical computing’s potential for increased bandwidth, which would enable increased throughput and enhanced security. Plasmonic devices have small footprints and interface with electronics easily, but these potential improvements are offset by the large device footprints of conventional signal regeneration schemes, since surface plasmon polaritons (SPPs) are incredibly lossy. As such, there is a need for novel regeneration schemes. The continuous, uniform, and unambiguous digital information encoding method is phase-shift-keying (PSK), so …

Kemar Hats Head Orientation Directivity, Samuel D. Bellows, Timothy W. Leishman

Directivity

This directivity data set for a KEMAR head head-and-torso simulator (HATS) includes head orientations in 14 directions in 5° steps starting from 0° to 40° and then in 10° steps from 40° to 90°. The full spherical measurements followed at an a = 0.97 m radius with the mouth aperture at the spherical center. The sampling density and distribution followed the AES 5° dual-equiangular sampling standard, omitting the south pole (θ = 180°). Thus, each spherical directivity assessment included 36 polar-angle θ samples and 72 azimuthal-angle ϕ samples. The presented data include 22 1/3-octave bands, ranging from 80 Hz …

Design Study For An Antenna Radar Cross Section Measurement Test Fixture, Wayne C. Kreimeyer

Theses and Dissertations

Military technology requires equipment to be undetected by adversaries. Stealth aircraft are designed to be undetected through electromagnetic means by minimizing a return signature called the RCS. Therefore, it is essential to understand how antennas, which are necessary for communication, affect the overall RCS of the aircraft. The antenna is measured in a compact RADAR range. The antenna needs a structure to support it, also referred to as a test fixture, that does not interfere with the measuring process of the antenna’s RCS. This thesis set out to get the lowest RCS possible of a test fixture by evaluating different …

Relative Magnetic Position And Rotation Sensor Assisted Dual-Foot Pedestrian Dead Reckoning, Jenario Y. Johnson

Theses and Dissertations

The use of wearable foot-based inertial measurement units (IMUs) incorporated in a navigation system can address the problem of single-person location tracking in situations and environments where GPS signals may be unavailable or inconsistent. This Pedestrian Dead Reckoning (PDR) approach enables standalone personal tracking. A notable solution involves using inertial measurement units (IMUs) in a filter to apply zero-velocity updates to a Kalman filter to get a position solution. This paper continues on the path of the former method by investigating the feasibility of PDR using a pair of low cost IMUs along with a pair of relative position and …

Evaluation Of The Armas-Som Framework With Real Data, Brandon M. Blakely

Theses and Dissertations

The ARMAS framework was created with the goal of developing a framework for all-source sensors which is able to combine detection, identification, calibration, model selection, and independent evaluation into a single system. Stable Observability Monitoring (SOM), augments the original ARMAS framework by enabling ARMAS to detect whether or not its Fault Detection and Exclusion (FDE) capabilities can be trusted and when additional sensor information is required to maintain resiliency. While previously tested with simulated sensor data, SOM has yet to be tested with real-life sensor data. Furthermore, ARMAS has only been tested with real-life GNSS data. This work expands on …

Global Sporadic-E Climatological Analysis Using Gps Radio Occultation And Ionosonde Data, Travis J. Hodos

Theses and Dissertations

A climatology of sporadic-E (Es) derived from a combined data set of GPS radio occultation (GPS-RO) and ground-based ionosonde soundings is presented for the period from September 2006 to February 2019. The ionosonde soundings were measured using the Lowell Digisonde International (LDI) Global Ionosphere Radio Observatory (GIRO) network consisting of 65 sites and 13,141,060 total soundings. The GPS-RO observations were taken aboard the Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) satellites and processed using two binary Es detection algorithms, totaling 9,072,922 occultations. The first algorithm is an S4 amplitude threshold calibrated to the occurrence of any blanketing Es …

Uav Positioning Data Determined Via Aruco Tags For Aircraft Surface Inspection, Caleb B. Schmidt

Theses and Dissertations

Aircraft are frequently inspected to ensure that military and civilian safety standards are adhered to. These inspections are performed pre- and post-flight and are currently performed by trained maintenance personnel. This work furthers the automation of aircraft surface inspection by using ArUco tags to determine the position of the UAV during aerial inspections. The ArUco tag based position data was then compared to a highly accurate infrared motion capture system to determine the viability of this for accurate positioning of the vehicle. This work includes flight experiments with two different UAVs to perform a system viability comparison.

Deep Learning Techniques To Estimate 3d Position In Stereoscopic Imagery, Jonathan I. Nicholson

Theses and Dissertations

Current AAR efforts utilize machine vision algorithms to estimate the pose of a receiver aircraft. However, these algorithms are dependent on several conditions such as the availability of precise 3D aircraft models; the accuracy of the pipeline significantly degrades in the absence of high-quality information given beforehand. We propose a deep learning architecture that estimates the 3D position of an object based on stereoscopic imagery. We investigate the use of both machine learning techniques and neural networks to directly regress the 3D position of the receiver aircraft. We present a new position estimation framework that is based on the differences …

Detection And Identification Of Cellphone Emitted Light Detection And Ranging Light In Security Camera Video Footage, Tristan V. Creek

Theses and Dissertations

The prevalence of Light Detection and Ranging (LiDAR) sensors in consumer cellphones has ushered in an era of access to technologies previously inaccessible to the average person. Although LiDAR sensors emit light in the infrared spectrum invisible to the naked eye, many security cameras possess the ability to capture infrared light reflected off surfaces to detect LiDAR light in an environment. Once detected and recorded to video footage, current classification methods fail to identify the LiDAR light. Therefore, this research develops a methodology to detect LiDAR light in security camera video footage and identify it as cellphone LiDAR. A proposed …