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.

Development of a Surface-Enhanced Raman Spectroscopy Method for the Detection of Benzodiazepines in Urine

Benzodiazepines are among the most prescribed compounds for anti-anxiety and are present in many toxicological screens. These drugs are also prominent in the commission of drug facilitated sexual assaults due their effects on the central nervous system. Due to their potency, a low dose of these compounds is often administered to victims; therefore, the target detection limit for these compounds in biological samples is 10 ng/mL. Currently these compounds are predominantly analyzed using immunoassay techniques; however more specific screening methods are needed. The goal of this dissertation was to develop a rapid, specific screening technique for benzodiazepines in urine samples utilizing surface-enhanced Raman spectroscopy (SERS), which has previously been shown be capable of to detect trace quantities of pharmaceutical compounds in aqueous solutions. Surface enhanced Raman spectroscopy has the advantage of overcoming the low sensitivity and fluorescence effects seen with conventional Raman spectroscopy. The spectra are obtained by applying an analyte onto a SERS-active metal substrate such as colloidal metal particles. SERS signals can be further increased with the addition of aggregate solutions. These agents cause the nanoparticles to amass and form hot-spots which increase the signal intensity. In this work, the colloidal particles are spherical gold nanoparticles in aqueous solution with an average size of approximately 30 nm. The optimum aggregating agent for the detection of benzodiazepines was determined to be 16.7 mM MgCl2, providing the highest signal intensities at the lowest drug concentrations with limits of detection between 0.5 and 127 ng/mL. A supported liquid extraction technique was utilized as a rapid clean extraction for benzodiazepines from urine at a pH of 5.0, allowing for clean extraction with limits of detection between 6 and 640 ng/mL. It was shown that at this pH other drugs that are prevalent in urine samples can be removed providing the selective detection of the benzodiazepine of interest. This technique has been shown to provide rapid (less than twenty minutes), sensitive, and specific detection of benzodiazepines at low concentrations in urine. It provides the forensic community with a sensitive and specific screening technique for the detection of benzodiazepines in drug facilitated assault cases.

Process Monitoring and Defect Detection of Additive Manufacturing Using Inline Coherent Imaging

With applications ranging from aerospace to biomedicine, additive manufacturing (AM) has been revolutionizing the manufacturing industry. The ability of additive techniques, such as selective laser melting (SLM), to create fully functional, geometrically complex, and unique parts out of high strength materials is of great interest. Unfortunately, despite numerous advantages afforded by this technology, its widespread adoption is hindered by a lack of on-line, real time feedback control and quality assurance techniques. In this thesis, inline coherent imaging (ICI), a broadband, spatially coherent imaging technique, is used to observe the SLM process in 15 - 45 $\mu m$ 316L stainless steel. Imaging of both single and multilayer builds is performed at a rate of 200 $kHz$, with a resolution of tens of microns, and a high dynamic range rendering it impervious to blinding from the process beam. This allows imaging before, during, and after laser processing to observe changes in the morphology and stability of the melt. Galvanometer-based scanning of the imaging beam relative to the process beam during the creation of single tracks is used to gain a unique perspective of the SLM process that has been so far unobservable by other monitoring techniques. Single track processing is also used to investigate the possibility of a preliminary feedback control parameter based on the process beam power, through imaging with both coaxial and 100 $\mu m$ offset alignment with respect to the process beam. The 100 $\mu m$ offset improved imaging by increasing the number of bright A-lines (i.e. with signal greater than the 10 $dB$ noise floor) by 300\%. The overlap between adjacent tracks in a single layer is imaged to detect characteristic fault signatures. Full multilayer builds are carried out and the resultant ICI images are used to detect defects in the finished part and improve upon the initial design of the build system. Damage to the recoater blade is assessed using powder layer scans acquired during a 3D build. The ability of ICI to monitor SLM processes at such high rates with high resolution offers extraordinary potential for future advances in on-line feedback control of additive manufacturing.

I/Q imbalance in two-way AF relaying: Performance analysis and detection mode switch

This paper studies the impact of in-phase and quadrature-phase imbalance (IQI) in two-way amplify-and-forward (AF) relaying systems. In particular, the effective signal-to-interference-plus-noise ratio (SINR) is derived for each source node, considering four different linear detection schemes, namely, uncompensated (Uncomp) scheme, maximal-ratio-combining (MRC), zero-forcing (ZF) and minimum mean-square error (MMSE) based schemes. For each proposed scheme, the outage probability (OP) is investigated over independent, non-identically distributed Nakagami-m fading channels, and exact closed-form expressions are derived for the first three schemes. Based on the closed-form OP expressions, an adaptive detection mode switching scheme is designed for minimizing the OP of both sources. An important observation is that, regardless of the channel conditions and transmit powers, the ZF-based scheme should always be selected if the target SINR is larger than 3 (4.77dB), while the MRC-based scheme should be avoided if the target SINR is larger than 0.38 (-4.20dB).

Enzyme-induced Aggregation of Plasmonic Nanoparticles for the Development of Label-free and Ultrasensitive Detection of Bacterial DNA

Sensitive detection of pathogens is critical to ensure the safety of food supplies and to prevent bacterial disease infection and outbreak at the first onset. While conventional techniques such as cell culture, ELISA, PCR, etc. have been used as the predominant detection workhorses, they are however limited by either time-consuming procedure, complicated sample pre-treatment, expensive analysis and operation, or inability to be implemented at point-of-care testing. Here, we present our recently developed assay exploiting enzyme-induced aggregation of plasmonic gold nanoparticles (AuNPs) for label-free and ultrasensitive detection of bacterial DNA. In the experiments, AuNPs are first functionalized with specific, single-stranded RNA probes so that they exhibit high stability in solution even under high electrolytic condition thus exhibiting red color. When bacterial DNA is present in a sample, a DNA-RNA heteroduplex will be formed and subsequently prone to the RNase H cleavage on the RNA probe, allowing the DNA to liberate and hybridize with another RNA strand. This continuously happens until all of the RNA strands are cleaved, leaving the nanoparticles ‘unprotected’. The addition of NaCl will cause the ‘unprotected’ nanoparticles to aggregate, initiating a colour change from red to blue. The reaction is performed in a multi-well plate format, and the distinct colour signal can be discriminated by naked eye or simple optical spectroscopy. As a result, bacterial DNA as low as pM could be unambiguously detected, suggesting that the enzyme-induced aggregation of AuNPs assay is very easy to perform and sensitive, it will significantly benefit to development of fast and ultrasensitive methods that can be used for disease detection and diagnosis.

A Reliable Downscaling of ECG Signals for the Detection of T wave Heterogeneity Features

In cardiovascular disease the definition and the detection of the ECG parameters related to repolarization dynamics in post MI patients is still a crucial unmet need. In addition, the use of a 3D sensor in the implantable medical devices would be a crucial mean in the assessment or prediction of Heart Failure status, but the inclusion of such feature is limited by hardware and firmware constraints. The aim of this thesis is the definition of a reliable surrogate of the 500 Hz ECG signal to reach the aforementioned objective. To evaluate the worsening of reliability due to sampling frequency reduction on delineation performance, the signals have been consecutively down sampled by a factor 2, 4, 8 thus obtaining the ECG signals sampled at 250, 125 and 62.5 Hz, respectively. The final goal is the feasibility assessment of the detection of the fiducial points in order to translate those parameters into meaningful clinical parameter for Heart Failure prediction, such as T waves intervals heterogeneity and variability of areas under T waves. An experimental setting for data collection on healthy volunteers has been set up at the Bakken Research Center in Maastricht. A 16 – channel ambulatory system, provided by TMSI, has recorded the standard 12 – Leads ECG, two 3D accelerometers and a respiration sensor. The collection platform has been set up by the TMSI property software Polybench, the data analysis of such signals has been performed with Matlab. The main results of this study show that the 125 Hz sampling rate has demonstrated to be a good candidate for a reliable detection of fiducial points. T wave intervals proved to be consistently stable, even at 62.5 Hz. Further studies would be needed to provide a better comparison between sampling at 250 Hz and 125 Hz for areas under the T waves.

Signal subspace identification in hyperspectral imagery

Terrestrial remote sensing imagery involves the acquisition of information from the Earth's surface without physical contact with the area under study. Among the remote sensing modalities, hyperspectral imaging has recently emerged as a powerful passive technology. This technology has been widely used in the fields of urban and regional planning, water resource management, environmental monitoring, food safety, counterfeit drugs detection, oil spill and other types of chemical contamination detection, biological hazards prevention, and target detection for military and security purposes [2-9]. Hyperspectral sensors sample the reflected solar radiation from the Earth surface in the portion of the spectrum extending from the visible region through the near-infrared and mid-infrared (wavelengths between 0.3 and 2.5 µm) in hundreds of narrow (of the order of 10 nm) contiguous bands [10]. This high spectral resolution can be used for object detection and for discriminating between different objects based on their spectral xharacteristics [6]. However, this huge spectral resolution yields large amounts of data to be processed. For example, the Airbone Visible/Infrared Imaging Spectrometer (AVIRIS) [11] collects a 512 (along track) X 614 (across track) X 224 (bands) X 12 (bits) data cube in 5 s, corresponding to about 140 MBs. Similar data collection ratios are achieved by other spectrometers [12]. Such huge data volumes put stringent requirements on communications, storage, and processing. The problem of signal sbspace identification of hyperspectral data represents a crucial first step in many hypersctral processing algorithms such as target detection, change detection, classification, and unmixing. The identification of this subspace enables a correct dimensionality reduction (DR) yelding gains in data storage and retrieval and in computational time and complexity. Additionally, DR may also improve algorithms performance since it reduce data dimensionality without losses in the useful signal components. The computation of statistical estimates is a relevant example of the advantages of DR, since the number of samples required to obtain accurate estimates increases drastically with the dimmensionality of the data (Hughes phnomenon) [13].

Anomaly Detection in Noisy Images

Finding rare events in multidimensional data is an important detection problem that has applications in many fields, such as risk estimation in insurance industry, finance, flood prediction, medical diagnosis, quality assurance, security, or safety in transportation. The occurrence of such anomalies is so infrequent that there is usually not enough training data to learn an accurate statistical model of the anomaly class. In some cases, such events may have never been observed, so the only information that is available is a set of normal samples and an assumed pairwise similarity function. Such metric may only be known up to a certain number of unspecified parameters, which would either need to be learned from training data, or fixed by a domain expert. Sometimes, the anomalous condition may be formulated algebraically, such as a measure exceeding a predefined threshold, but nuisance variables may complicate the estimation of such a measure. Change detection methods used in time series analysis are not easily extendable to the multidimensional case, where discontinuities are not localized to a single point. On the other hand, in higher dimensions, data exhibits more complex interdependencies, and there is redundancy that could be exploited to adaptively model the normal data. In the first part of this dissertation, we review the theoretical framework for anomaly detection in images and previous anomaly detection work done in the context of crack detection and detection of anomalous components in railway tracks. In the second part, we propose new anomaly detection algorithms. The fact that curvilinear discontinuities in images are sparse with respect to the frame of shearlets, allows us to pose this anomaly detection problem as basis pursuit optimization. Therefore, we pose the problem of detecting curvilinear anomalies in noisy textured images as a blind source separation problem under sparsity constraints, and propose an iterative shrinkage algorithm to solve it. Taking advantage of the parallel nature of this algorithm, we describe how this method can be accelerated using graphical processing units (GPU). Then, we propose a new method for finding defective components on railway tracks using cameras mounted on a train. We describe how to extract features and use a combination of classifiers to solve this problem. Then, we scale anomaly detection to bigger datasets with complex interdependencies. We show that the anomaly detection problem naturally fits in the multitask learning framework. The first task consists of learning a compact representation of the good samples, while the second task consists of learning the anomaly detector. Using deep convolutional neural networks, we show that it is possible to train a deep model with a limited number of anomalous examples. In sequential detection problems, the presence of time-variant nuisance parameters affect the detection performance. In the last part of this dissertation, we present a method for adaptively estimating the threshold of sequential detectors using Extreme Value Theory on a Bayesian framework. Finally, conclusions on the results obtained are provided, followed by a discussion of possible future work.

MOTION - A New On-line Model for Urban Traffic Signal Control.

This paper presents MOTION, a modular on-line model for urban traffic signal control. It consists of a network and a local level and builds on enhanced traffic state estimation. Special consideration is given to the prioritization of public transit. MOTION provides possibilities for the interaction with integrated urban management systems.

'Danger Theory: The Link between AIS and IDS?'

Abstract We present ideas about creating a next generation Intrusion Detection System (IDS) based on the latest immunological theories. The central challenge with computer security is determining the difference between normal and potentially harmful activity. For half a century, developers have protected their systems by coding rules that identify and block specific events. However, the nature of current and future threats in conjunction with ever larger IT systems urgently requires the development of automated and adaptive defensive tools. A promising solution is emerging in the form of Artificial Immune Systems (AIS): The Human Immune System (HIS) can detect and defend against harmful and previously unseen invaders, so can we not build a similar Intrusion Detection System (IDS) for our computers? Presumably, those systems would then have the same beneficial properties as HIS like error tolerance, adaptation and self-monitoring. Current AIS have been successful on test systems, but the algorithms rely on self-nonself discrimination, as stipulated in classical immunology. However, immunologist are increasingly finding fault with traditional self-nonself thinking and a new 'Danger Theory' (DT) is emerging. This new theory suggests that the immune system reacts to threats based on the correlation of various (danger) signals and it provides a method of 'grounding' the immune response, i.e. linking it directly to the attacker. Little is currently understood of the precise nature and correlation of these signals and the theory is a topic of hot debate. It is the aim of this research to investigate this correlation and to translate the DT into the realms of computer security, thereby creating AIS that are no longer limited by self-nonself discrimination. It should be noted that we do not intend to defend this controversial theory per se, although as a deliverable this project will add to the body of knowledge in this area. Rather we are interested in its merits for scaling up AIS applications by overcoming self-nonself discrimination problems.

Detecting Danger: Applying a Novel Immunological Concept to Intrusion Detection Systems'

INTRODUCTION In recent years computer systems have become increasingly complex and consequently the challenge of protecting these systems has become increasingly difficult. Various techniques have been implemented to counteract the misuse of computer systems in the form of firewalls, antivirus software and intrusion detection systems. The complexity of networks and dynamic nature of computer systems leaves current methods with significant room for improvement. Computer scientists have recently drawn inspiration from mechanisms found in biological systems and, in the context of computer security, have focused on the human immune system (HIS). The human immune system provides an example of a robust, distributed system that provides a high level of protection from constant attacks. By examining the precise mechanisms of the human immune system, it is hoped the paradigm will improve the performance of real intrusion detection systems. This paper presents an introduction to recent developments in the field of immunology. It discusses the incorporation of a novel immunological paradigm, Danger Theory, and how this concept is inspiring artificial immune systems (AIS). Applications within the context of computer security are outlined drawing direct reference to the underlying principles of Danger Theory and finally, the current state of intrusion detection systems is discussed and improvements suggested.

Precoder Detection for Cooperative Decode-and-Forward Relaying in OFDMA Systems

We consider an LTE network where a secondary user acts as a relay, transmitting data to the primary user using a decode-and-forward mechanism, transparent to the base-station (eNodeB). Clearly, the relay can decode symbols more reliably if the employed precoder matrix indicators (PMIs) are known. However, for closed loop spatial multiplexing (CLSM) transmit mode, this information is not always embedded in the downlink signal, leading to a need for effective methods to determine the PMI. In this thesis, we consider 2x2 MIMO and 4x4 MIMO downlink channels corresponding to CLSM and formulate two techniques to estimate the PMI at the relay using a hypothesis testing framework. We evaluate their performance via simulations for various ITU channel models over a range of SNR and for different channel quality indicators (CQIs). We compare them to the case when the true PMI is known at the relay and show that the performance of the proposed schemes are within 2 dB at 10% block error rate (BLER) in almost all scenarios. Furthermore, the techniques add minimal computational overhead over existent receiver structure. Finally, we also identify scenarios when using the proposed precoder detection algorithms in conjunction with the cooperative decode-and-forward relaying mechanism benefits the PUE and improves the BLER performance for the PUE. Therefore, we conclude from this that the proposed algorithms as well as the cooperative relaying mechanism at the CMR can be gainfully employed in a variety of real-life scenarios in LTE networks.