Jack McLaughlin

Non-stationary signal classification using joint frequency analysis

Time-varying short-term spectral estimates have been successfully applied in many classification tasks. However, they are still insufficient for many non-stationary signals where time-varying information is useful. We propose to improve the deficiencies of current short-term feature analysis by adding information to describe the time-varying behavior of the signals. Our proposed method, which is motivated by the human auditory system, can be applied to several non-stationary signal types. Real world communication signals were used for experimental verification. These experimental results, assessed with a conventional probabilistic classifier, showed significant improvement when the new features were added to short-term spectral estimates.

Improving target tracking performance by incorporating classification information

This paper presents approaches for incorporating classification information into target tracking algorithms, specifically in a multistatic active sonar context. In addition, this paper describes the framework designed for simulation and classification of return time series from simulated targets and clutter in a realistic underwater environment. The simulated target and clutter returns are integrated into an existing contact-based tracking dataset (TNO Blind dataset) for which time series are unavailable. Simulations compare the integrating classification of contacts at different stages of tracking algorithms. Results show improvements in some tracking metrics with no degradation of the others.

Using speech technology to enhance isotope ID and classification

Scientific advances are often made when researchers identify mathematical or physical commonalities between different fields and are able to apply mature techniques or algorithms developed in one field to another field which shares some of the same challenges. The authors of this paper have identified similarities between the unsolved problems faced in gamma-spectroscopy for automated radioisotope identification and the challenges of the much larger body of research in speech processing. In this paper we describe such commonalities and use them as a motivation for a preliminary investigation of the applicability of speech processing methods to gamma-ray spectra. This approach enables the development of proof-of-concept isotope classifiers, whose performance is presented for both simulated and field-collected gamma-ray spectra.

Speaker detection and tracking for telephone transactions

As ever greater numbers of telephone transactions are being conducted solely between a caller and an automated answering system, the need increases for software which can automatically identify and authenticate these callers without the need for an onerous speaker enrollment process. In this paper we introduce and investigate a novel speaker detection and tracking (SDT) technique, which dynamically merges the traditional enrollment and recognition phases of the static speaker recognition task. In this speaker recognition application, no prior speaker models exist and the goal is to detect and model new speakers as they call into the system while also recognizing utterances from previously modeled callers. New speakers are added to the enrolled set of speakers and speech from speakers in the currently enrolled set is used to update models. We describe a system based on a GMM speaker identification (SID) system and develop a new measure to evaluate the performance of the system on the SDT task. Results for both static, open-set detection and the SDT task are presented using a portion of the Switchboard corpus of telephone speech communications. Static open-set detection produces an equal error rate of about 5%. As expected, performance for SDT is quite varied, depending greatly on the speaker set and ordering of the test sequence. These initial results, however, are quite promising and point to potential areas in which to improve the system performance.

A technique for adjusting Gaussian mixture model weights that improves speaker identification performance in the presence of phonemic train/test mismatch.

Speaker identification is complicated by cases where training material is phonemically deficient. Misclassifications can result either because subsequent test material from that speaker contains primarily the phonemes missing from the training data or because that test material is phonemically most consistent with another talker’s model. This situation can arise in any dialog where, for reasons of brevity and clarity, conventions must be imposed on phraseology. We present here a technique for detecting phonemic deficiencies in a speaker model, and then correcting that model to partially compensate for the biased training data. This technique relies upon a specially constructed universal background model (UBM) from which speaker models are adapted. This UBM is formed by weighting several dozen phoneme GMMs using EM training. As a result, each Gaussian component of the UBM (and of the resulting speaker models) corresponds to a specific phoneme. Analysis of the speaker model weights reveals whether the train...

Application of low‐frequency methods for estimating object size.

Classification of submerged objects has traditionally been performed using high frequency sonars and imaging techniques. While this permits fine matching of target templates to images acquired in the field, HF methods are necessarily limited in range due to absorption of sound by the water. LF sonars, while offering increased detection range, come with some significant challenges related to the limited bandwidth available. Nonetheless, we show that it is feasible to estimate object size using nonimaging techniques. There are a number of low‐frequency phenomena that can be exploited to this end. Among these are edge diffraction in which sharply angled facets of objects (“edges”) act like independent, radiating point sources, and helical waves, which can be set up in cylindrical objects. We show that with appropriate postprocessing of these returns, object edges can be localized thus allowing object extent to be assessed. In this paper, we describe our processing system, and then give results when this syst...

Modeling acoustic target response by component

The scattered acoustic response of underwater objects due to active interrogation has been studied for decades for use in detection and classification applications. As a means of detection, fielded applications date back nearly a hundred years. However, use of responses for robust automated classification has lagged behind, particularly when the internal structure of the objects is of key importance and when the objects may be partially or fully buried. Analytic solutions for simple geometries have provided much understanding of certain physical mechanisms, but transfer to complex structures of practical importance has proven difficult. In recent decades, finite element (FE) modeling has provided a method of accurate simulation of many structures previously considered intractable. However, simulation of such complex objects produces equally complex returns, with the result that the models are often simply considered as a “black box” where the physical interpretation of the response components is tenuous at best. Thus the state of the art is still short of a method for development of robust classification systems for complex objects based on the physics of the objects of interest and the varied conditions under which they may be found. This paper introduces an effort to use FE techniques to simulate individual components of a return by “turning off” most aspects of the physics and allowing the researcher to isolate one mechanism at a time. The goal is a true physical understanding of the complete response, a physically justifiable feature set for classification, and a much simpler path to environmental robustness.

Improving anti‐submarine warfare tracking performance by incorporating classification information.

Prior research has shown that more accurate classification of contact amplitudes can improve tracking performance. In this work, we augment standard, contact‐based trackers with a classifier run on features of the received time series from which the contacts were extracted. Ground truth information from benchmark datasets is used in a flexible simulation framework built around the sonar simulation toolset (SST) to generate simulated target and clutter returns. The simulated returns are used as input to the classifier and the contacts from the benchmark datasets as input to the tracker. The classification information provides an additional input to the association step in the probabilistic data association (PDA) and joint PDA trackers, and to the probability of target for each contact in the Bayesian tracker. The results show that even relatively poor classification can make a noticeable improvement in tracking performance. [This work was funded by the U.S. Office of Naval Research, Contract No. N00014‐01‐G‐0460, Delivery Order #36.]

Finding physically motivated classification features using finite element models.

Finite element models (FEMs) for simple shapes such as spheres, cylinders and pipes can be used to uncover potential features for classification by sonar systems. Analyses were conducted of the simulated, bistatic returns over a wide range of aspects using FEMs for a solid steel cylinder and a cement pipe all in free space. Salient characteristics noted in this way were then sought in field data collected from the same or similar objects during the SAX04 experiment and two other data collections conducted in a test pond at the Naval Surface Warfare Center in Panama City, Florida. Characteristics robust enough to appear in both simulated data and field data are candidates to serve as classification features for similarly shaped objects. It is found that FEMs can provide important pointers to classification features that are closely tied to the physics of reflection. As such, the variability of these features in the face of object burial and multipath can be more easily assessed than statistical spectral fe...