Detection and Classification of Whale Acoustic Signals

This dissertation focuses on two vital challenges in relation to whale acoustic signals: detection and classification.

In detection, we evaluated the influence of the uncertain ocean environment on the spectrogram-based detector, and derived the likelihood ratio of the proposed Short Time Fourier Transform detector. Experimental results showed that the proposed detector outperforms detectors based on the spectrogram. The proposed detector is more sensitive to environmental changes because it includes phase information.

In classification, our focus is on finding a robust and sparse representation of whale vocalizations. Because whale vocalizations can be modeled as polynomial phase signals, we can represent the whale calls by their polynomial phase coefficients. In this dissertation, we used the Weyl transform to capture chirp rate information, and used a two dimensional feature set to represent whale vocalizations globally. Experimental results showed that our Weyl feature set outperforms chirplet coefficients and MFCC (Mel Frequency Cepstral Coefficients) when applied to our collected data.

Since whale vocalizations can be represented by polynomial phase coefficients, it is plausible that the signals lie on a manifold parameterized by these coefficients. We also studied the intrinsic structure of high dimensional whale data by exploiting its geometry. Experimental results showed that nonlinear mappings such as Laplacian Eigenmap and ISOMAP outperform linear mappings such as PCA and MDS, suggesting that the whale acoustic data is nonlinear.

We also explored deep learning algorithms on whale acoustic data. We built each layer as convolutions with either a PCA filter bank (PCANet) or a DCT filter bank (DCTNet). With the DCT filter bank, each layer has different a time-frequency scale representation, and from this, one can extract different physical information. Experimental results showed that our PCANet and DCTNet achieve high classification rate on the whale vocalization data set. The word error rate of the DCTNet feature is similar to the MFSC in speech recognition tasks, suggesting that the convolutional network is able to reveal acoustic content of speech signals.

Coding Strategies and Implementations of Compressive Sensing

This dissertation studies the coding strategies of computational imaging to overcome the limitation of conventional sensing techniques. The information capacity of conventional sensing is limited by the physical properties of optics, such as aperture size, detector pixels, quantum efficiency, and sampling rate. These parameters determine the spatial, depth, spectral, temporal, and polarization sensitivity of each imager. To increase sensitivity in any dimension can significantly compromise the others.

This research implements various coding strategies subject to optical multidimensional imaging and acoustic sensing in order to extend their sensing abilities. The proposed coding strategies combine hardware modification and signal processing to exploiting bandwidth and sensitivity from conventional sensors. We discuss the hardware architecture, compression strategies, sensing process modeling, and reconstruction algorithm of each sensing system.

Optical multidimensional imaging measures three or more dimensional information of the optical signal. Traditional multidimensional imagers acquire extra dimensional information at the cost of degrading temporal or spatial resolution. Compressive multidimensional imaging multiplexes the transverse spatial, spectral, temporal, and polarization information on a two-dimensional (2D) detector. The corresponding spectral, temporal and polarization coding strategies adapt optics, electronic devices, and designed modulation techniques for multiplex measurement. This computational imaging technique provides multispectral, temporal super-resolution, and polarization imaging abilities with minimal loss in spatial resolution and noise level while maintaining or gaining higher temporal resolution. The experimental results prove that the appropriate coding strategies may improve hundreds times more sensing capacity.

Human auditory system has the astonishing ability in localizing, tracking, and filtering the selected sound sources or information from a noisy environment. Using engineering efforts to accomplish the same task usually requires multiple detectors, advanced computational algorithms, or artificial intelligence systems. Compressive acoustic sensing incorporates acoustic metamaterials in compressive sensing theory to emulate the abilities of sound localization and selective attention. This research investigates and optimizes the sensing capacity and the spatial sensitivity of the acoustic sensor. The well-modeled acoustic sensor allows localizing multiple speakers in both stationary and dynamic auditory scene; and distinguishing mixed conversations from independent sources with high audio recognition rate.

Integrated System Technologies for Modular Trapped Ion Quantum Information Processing

Although trapped ion technology is well-suited for quantum information science, scalability of the system remains one of the main challenges. One of the challenges associated with scaling the ion trap quantum computer is the ability to individually manipulate the increasing number of qubits. Using micro-mirrors fabricated with micro-electromechanical systems (MEMS) technology, laser beams are focused on individual ions in a linear chain and steer the focal point in two dimensions. Multiple single qubit gates are demonstrated on trapped 171Yb+ qubits and the gate performance is characterized using quantum state tomography. The system features negligible crosstalk to neighboring ions (< 3e-4), and switching speeds comparable to typical single qubit gate times (< 2 us). In a separate experiment, photons scattered from the 171Yb+ ion are coupled into an optical fiber with 63% efficiency using a high numerical aperture lens (0.6 NA). The coupled photons are directed to superconducting nanowire single photon detectors (SNSPD), which provide a higher detector efficiency (69%) compared to traditional photomultiplier tubes (35%). The total system photon collection efficiency is increased from 2.2% to 3.4%, which allows for fast state detection of the qubit. For a detection beam intensity of 11 mW/cm2, the average detection time is 23.7 us with 99.885(7)% detection fidelity. The technologies demonstrated in this thesis can be integrated to form a single quantum register with all of the necessary resources to perform local gates as well as high fidelity readout and provide a photon link to other systems.

Learning and Exploiting Camera Geometry for Computer Vision

This work explores the use of statistical methods in describing and estimating camera poses, as well as the information feedback loop between camera pose and object detection. Surging development in robotics and computer vision has pushed the need for algorithms that infer, understand, and utilize information about the position and orientation of the sensor platforms when observing and/or interacting with their environment.

The first contribution of this thesis is the development of a set of statistical tools for representing and estimating the uncertainty in object poses. A distribution for representing the joint uncertainty over multiple object positions and orientations is described, called the mirrored normal-Bingham distribution. This distribution generalizes both the normal distribution in Euclidean space, and the Bingham distribution on the unit hypersphere. It is shown to inherit many of the convenient properties of these special cases: it is the maximum-entropy distribution with fixed second moment, and there is a generalized Laplace approximation whose result is the mirrored normal-Bingham distribution. This distribution and approximation method are demonstrated by deriving the analytical approximation to the wrapped-normal distribution. Further, it is shown how these tools can be used to represent the uncertainty in the result of a bundle adjustment problem.

Another application of these methods is illustrated as part of a novel camera pose estimation algorithm based on object detections. The autocalibration task is formulated as a bundle adjustment problem using prior distributions over the 3D points to enforce the objects' structure and their relationship with the scene geometry. This framework is very flexible and enables the use of off-the-shelf computational tools to solve specialized autocalibration problems. Its performance is evaluated using a pedestrian detector to provide head and foot location observations, and it proves much faster and potentially more accurate than existing methods.

Finally, the information feedback loop between object detection and camera pose estimation is closed by utilizing camera pose information to improve object detection in scenarios with significant perspective warping. Methods are presented that allow the inverse perspective mapping traditionally applied to images to be applied instead to features computed from those images. For the special case of HOG-like features, which are used by many modern object detection systems, these methods are shown to provide substantial performance benefits over unadapted detectors while achieving real-time frame rates, orders of magnitude faster than comparable image warping methods.

The statistical tools and algorithms presented here are especially promising for mobile cameras, providing the ability to autocalibrate and adapt to the camera pose in real time. In addition, these methods have wide-ranging potential applications in diverse areas of computer vision, robotics, and imaging.

Effective Dose Estimation in Fast-kV Switch Dual Energy Computed Tomography

Purpose

The objective of our study was to test a new approach to approximating organ dose by using the effective energy of the combined 80kV/140kV beam used in fast kV switch dual-energy (DE) computed tomography (CT). The two primary focuses of the study were to first validate experimentally the dose equivalency between MOSFET and ion chamber (as a gold standard) in a fast kV switch DE environment, and secondly to estimate effective dose (ED) of DECT scans using MOSFET detectors and an anthropomorphic phantom.

Materials and Methods

A GE Discovery 750 CT scanner was employed using a fast-kV switch abdomen/pelvis protocol alternating between 80 kV and 140 kV. The specific aims of our study were to (1) Characterize the effective energy of the dual energy environment; (2) Estimate the f-factor for soft tissue; (3) Calibrate the MOSFET detectors using a beam with effective energy equal to the combined DE environment; (4) Validate our calibration by using MOSFET detectors and ion chamber to measure dose at the center of a CTDI body phantom; (5) Measure ED for an abdomen/pelvis scan using an anthropomorphic phantom and applying ICRP 103 tissue weighting factors; and (6) Estimate ED using AAPM Dose Length Product (DLP) method. The effective energy of the combined beam was calculated by measuring dose with an ion chamber under varying thicknesses of aluminum to determine half-value layer (HVL).

Results

The effective energy of the combined dual-energy beams was found to be 42.8 kV. After calibration, tissue dose in the center of the CTDI body phantom was measured at 1.71 ± 0.01 cGy using an ion chamber, and 1.73±0.04 and 1.69±0.09 using two separate MOSFET detectors. This result showed a -0.93% and 1.40 % difference, respectively, between ion chamber and MOSFET. ED from the dual-energy scan was calculated as 16.49 ± 0.04 mSv by the MOSFET method and 14.62 mSv by the DLP method.

X-Ray Diffraction Study of Boron Produced by Pyrolysis of Boron Tribromide

The goal of this research was to determine the composition of boron deposits produced by pyrolysis of boron tribromide, and to use the results to (a) determine the experimental conditions (reaction temperature, etc.) necessary to produce alpha-rhombohedral boron and (b) guide the development/refinement of the pyrolysis experiments such that large, high purity crystals of alpha-rhombohedral boron can be produced with consistency. Developing a method for producing large, high purity alpha-rhombohedral boron crystals is of interest because such crystals could potentially be used to achieve an alpha-rhombohedral boron based neutron detector design (a solid-state detector) that could serve as an alternative to existing neutron detector technologies. The supply of neutron detectors in the United States has been hampered for a number of years due to the current shortage of helium-3 (a gas used in many existing neutron detector technologies); the development of alternative neutron detector technology such as an alpha-rhombohedral boron based detector would help provide a more sustainable supply of neutron detectors in this country. In addition, the prospect/concept of an alpha-rhombohedral boron based neutron detector is attractive because it offers the possibility of achieving a design that is smaller, longer life, less power consuming, and potentially more sensitive than existing neutron detectors. The main difficulty associated with creating an alpha-rhombohedral boron based neutron detector is that producing large, high purity crystals of alpha-rhombohedral boron is extremely challenging. Past researchers have successfully made alpha-rhombohedral boron via a number of methods, but no one has developed a method for consistently producing large, high purity crystals. Alpha-rhombohedral boron is difficult to make because it is only stable at temperatures below around 1100-1200 °C, its formation is very sensitive to impurities, and the conditions necessary for its formation are not fully understood or agreed upon in the literature. In this research, the method of pyrolysis of boron tribromide (hydrogen reduction of boron tribromide) was used to deposit boron on a tantalum filament. The goal was to refine this method, or potentially use it in combination with a second method (amorphous boron crystallization), to the point where it is possible to grow large, high purity alpha-rhombohedral boron crystals with consistency. A pyrolysis apparatus was designed and built, and a number of trials were run to determine the conditions (reaction temperature, etc.) necessary for alpha-rhombohedral boron production. This work was focused on the x-ray diffraction analysis of the boron deposits; x-ray diffraction was performed on a number of samples to determine the types of boron (and other compounds) formed in each trial and to guide the choices of test conditions for subsequent trials. It was found that at low reaction temperatures (in the range of around 830-950 °C), amorphous boron was the primary form of boron produced. Reaction temperatures in the range of around 950-1000 °C yielded various combinations of crystalline boron and amorphous boron. In the first trial performed at a temperature of 950 °C, a mix of amorphous boron and alpha-rhombohedral boron was formed. Using a scanning electron microscope, it was possible to see small alpha-rhombohedral boron crystals (on the order of ~1 micron in size) embedded in the surface of the deposit. In subsequent trials carried out at reaction temperatures in the range of 950 °C – 1000 °C, it was found that various combinations of alpha-rhombohedral boron, beta-rhombohedral boron, and amorphous boron were produced; the results tended to be unpredictable (alpha-rhombohedral boron was not produced in every trial), and the factors leading to success/failure were difficult to pinpoint. These results illustrate how sensitive of a process producing alpha-rhombohedral boron can be, and indicate that further improvements to the test apparatus and test conditions (for example, higher purity/cleanliness) may be necessary to optimize the boron deposition. Although alpha-rhombohedral boron crystals of large size were not achieved, this research was successful in (a) developing a pyrolysis apparatus and test procedure that can serve as a platform for future testing, (b) determining reaction temperatures at which alpha-rhombohedral boron can form, and (c) developing a consistent process for analyzing the boron deposits and determining their composition. Further experimentation is necessary to achieve a pyrolysis apparatus and test procedure that can yield large alpha-rhombohedral boron crystals with consistency.

Preparation of Boron by Pyrolytic Decomposition of Boron Tribromide

The national shortage of helium-3 has made it critical to develop an alternative to helium-3 neutron detectors. Boron-10, if it could be produced in macroscopic alpha-rhombohedral crystalline form, would be a viable alternative to helium-3. This work has determined the critical parameters needed for the preparation of alpha-rhombohedral boron by the pyrolytic decomposition of boron tribromide on tantalum wire. The primary parameters that must be met are wire temperature and feedstock purity. The minimum purity level for boron tribromide was determined to be 99.999% and it has been found that alpha-rhombohedral boron cannot be produced using 99.99% boron tribromide. The decomposition temperature was experimentally tested between 830°C and 1000°C. Alpha-rhombohedral boron was found at temperatures between 950°C and 1000°C using 99.999% pure boron tribromide.

Measuring the $^{127}$I, $\nu_{e}$ Charged-Current Cross-Section of NaI[Tl]

The primary objective of this experiment is to measure the cross-section of $\nu_{e}$ charged-current neutrino interactions on $^{127}$I. To measure this interaction, an array of twenty-four, 7.7 kg sodium iodide (NaI[Tl]) scintillating detectors will be deployed to the Spallation Neutron Source at Oak Ridge National Laboratory. The design of the detector array is presented here along with preliminary characterization and background measurements conducted at Duke University.