Publications

Export 5 results:
Sort by: Author [ Title  (Desc)] Type Year
A B [C] D E F G H I J K L M N O P Q R S T U V W X Y Z   [Show ALL]
C
Soldevilla, MS, McKenna ME, Wiggins SM, Shadwick RE, Cranford TW, Hildebrand JA.  2005.  Cuvier's beaked whale (Ziphius cavirostris) head tissues: physical properties and CT imaging. Journal of Experimental Biology. 208:2319-2332.   10.1242/jeb.01624   AbstractWebsite

Tissue physical properties from a Clavier's beaked whale (Ziphius cavirostris) neonate head are reported and compared with computed tomography (CT) X-ray imaging. Physical properties measured include longitudinal sound velocity, density, elastic modulus and hysteresis. Tissues were classified by type as follows: mandibular acoustic fat, mandibular blubber, forehead acoustic fat (melon), forehead blubber, muscle and connective tissue. Results show that each class of tissues has unique, co-varying physical properties. The mandibular acoustic fats had minimal values for sound speed (1350 +/- 10.6 m s(-1)) and mass density (890 +/- 23 kg m(-3)). These values increased through mandibular blubber (1376 +/- 13 m s(-1), 919 +/- 13 kg m(-3)), melon (1382 +/- 23m s(-1), 937 +/- 17 kg m(-3)), forehead blubber (1401 +/- 7.8 m s(-1), 935 +/- 25 kg m(-3)) and muscle (1517 +/- 46.8 m s(-1), 993 +/- 58 kg m(-3)). Connective tissue had the greatest I mean sound speed and density (1628 +/- 48.7 m s(-1)

Roch, MA, Stinner-Sloan J, Baumann-Pickering S, Wiggins SM.  2015.  Compensating for the effects of site and equipment variation on delphinid species identification from their echolocation clicks. Journal of the Acoustical Society of America. 137:22-29.   10.1121/1.4904507   AbstractWebsite

A concern for applications of machine learning techniques to bioacoustics is whether or not classifiers learn the categories for which they were trained. Unfortunately, information such as characteristics of specific recording equipment or noise environments can also be learned. This question is examined in the context of identifying delphinid species by their echolocation clicks. To reduce the ambiguity between species classification performance and other confounding factors, species whose clicks can be readily distinguished were used in this study: Pacific white-sided and Risso's dolphins. A subset of data from autonomous acoustic recorders located at seven sites in the Southern California Bight collected between 2006 and 2012 was selected. Cepstral-based features were extracted for each echolocation click and Gaussian mixture models were used to classify groups of 100 clicks. One hundred Monte-Carlo three-fold experiments were conducted to examine classification performance where fold composition was determined by acoustic encounter, recorder characteristics, or recording site. The error rate increased from 6.1% when grouped by acoustic encounter to 18.1%, 46.2%, and 33.2% for grouping by equipment, equipment category, and site, respectively. A noise compensation technique reduced error for these grouping schemes to 2.7%, 4.4%, 6.7%, and 11.4%, respectively, a reduction in error rate of 56%-86%. (C) 2015 Acoustical Society of America.

Hildebrand, JA, Wiggins SM, Henkart PC, Conyers LB.  2002.  Comparison of seismic reflection and ground-penetrating radar imaging at the controlled archaeological test site, Champaign, Illinois. Archaeological Prospection. 9:9-21.   10.1002/arp.177   Abstract

Shallow seismic reflection and ground-penetrating radar images were collected at a replicated burial mound in the Controlled Archaeological Test Site (CATS) in Champaign, Illinois. The CATS mound contains a pig burial within a wood-lined crypt at a depth of 1.6–2.4 m. Seismic reflection data were collected from two different energy sources: a small (0.5 kg) hammer for an impulsive source, and a vibrator for a frequency swept source. Seismic data were collected at densely spaced points (5 cm) along a line of 48 geophone receivers. These data were stacked in a common mid-point gather, band-pass filtered, and processed with frequency–wavenumber migration. The seismic image produced by the hammer source was dominated by bodywaves at 120 Hz, whereas the vibrator source image was dominated by surface waves at 70 Hz. Both seismic sources revealed clear reflections from the burial crypt, and placed the top of the crypt at the correct depth with a seismic velocity of 120 m s 1. The bottom of the crypt was poorly defined by the seismic data owing to multiple reflections within the crypt. The vibrator source also revealed a highfrequency (360 Hz) reflector at 2.7 m depth within the mound, perhaps due to a resonant cavity within the pig’s body. Single channel ground-penetrating radar data were processed with the same approach, including band-pass filtering and migration. The radar data reveal clear reflections from the burial crypt. Extremely fast radar velocities (260 mm ns 1) are required in the upper portion of the burial mound to place the top of the crypt at its correct depth. The bottom of the crypt was well defined by ground-penetrating radar, and was located accurately with respect to the top of the crypt with a moderate radar velocity (170 mm ns 1). The application of both seismic reflection and ground-penetrating radar to the same site may be beneficial for improved understanding of their abilities for shallow subsurface imaging. Copyright  2002 John Wiley & Sons, Ltd.

Roch, M. A., Soldevilla MS, Hoenigman R, Wiggins SM, Hildebrand J.  2008.  Comparison of machine learning techniques for the classification of echolocation clicks from three species of Odontocetes. Canadian Acoustics. 36:41-47. AbstractWebsite

A species detectorclassifier is presented which decides whether or not short groups of clicks are produced by one or more individuals from the following species: Blainville’s beaked whales, short- finned pilot whales, and Risso’s dolphins. The system locates individual clicks using the Teager energy operator and then constructs feature vectors for these clicks using cepstral analysis. Two different types of detectors confirm or reject the presence of each species. Gaussian mixture models (GMMs) are used to model time series independent characteristics of the species feature vector distributions. Support vector machines (SVMs) are used to model the boundaries between each species’ feature distribution and that of other species. Detection error tradeoff curves for all three species are shown with the following equal error rates: Blainville’s beaked whales (GMM 3.32%/SVM 5.54%), pilot whales (GMM 16.18%/SVM 15.00%), and Risso’s dolphins (GMM 0.03%/SVM 0.70%).

Soldevilla, MS, Henderson EE, Campbell GS, Wiggins SM, Hildebrand JA, Roch MA.  2008.  Classification of Risso's and Pacific white-sided dolphins using spectral properties of echolocation clicks. The Journal of the Acoustical Society of America. 124:609-624.: ASA   10.1121/1.2932059   AbstractWebsite

The spectral and temporal properties of echolocation clicks and the use of clicks for species classification are investigated for five species of free-ranging dolphins found offshore of southern California: short-beaked common (Delphinus delphis), long-beaked common (D. capensis), Risso's (Grampus griseus), Pacific white-sided (Lagenorhynchus obliquidens), and bottlenose (Tursiops truncatus) dolphins. Spectral properties are compared among the five species and unique spectral peak and notch patterns are described for two species. The spectral peak mean values from Pacific white-sided dolphin clicks are 22.2, 26.6, 33.7, and 37.3 kHz and from Risso's dolphins are 22.4, 25.5, 30.5, and 38.8 kHz. The spectral notch mean values from Pacific white-sided dolphin clicks are 19.0, 24.5, and 29.7 kHz and from Risso's dolphins are 19.6, 27.7, and 35.9 kHz. Analysis of variance analyses indicate that spectral peaks and notches within the frequency band 24–35 kHz are distinct between the two species and exhibit low variation within each species. Post hoc tests divide Pacific white-sided dolphin recordings into two distinct subsets containing different click types, which are hypothesized to represent the different populations that occur within the region. Bottlenose and common dolphin clicks do not show consistent patterns of spectral peaks or notches within the frequency band examined (1–100 kHz).