Match!

Multifrequency species classification of acoustic-trawl survey data using semi-supervised learning with class discovery

Published on Apr 1, 2011in Journal of the Acoustical Society of America1.819
· DOI :10.1121/1.3678685
Mathieu Woillez12
Estimated H-index: 12
,
Patrick H. Ressler17
Estimated H-index: 17
+ 1 AuthorsJohn K. Horne25
Estimated H-index: 25
Abstract
Acoustic surveys often use multifrequency backscatter to estimate fish and plankton abundance. Direct samples are used to validate species classification of acoustic backscatter, but samples may be sparse or unavailable. A generalized Gaussian mixture model was developed to classify multifrequency acoustic backscatter when not all species classes are known. The classification, based on semi-supervised learning with class discovery, was applied to data collected in the eastern Bering Sea during summers 2004, 2007, and 2008. Walleye pollock, euphausiids, and two other major classes occurring in the upper water column were identified.
  • References (11)
  • Citations (14)
📖 Papers frequently viewed together
80 Citations
152 Citations
63 Citations
78% of Scinapse members use related papers. After signing in, all features are FREE.
References11
Newest
80 CitationsSource
63 CitationsSource
Fernandes, P. G. 2009. Classification trees for species identification of fish-school echotraces. - ICES Journal of Marine Science, 66: 1073-1080.Acoustic surveys provide valuable information on the abundance and distribution of many fish species, but are particularly effective for schooling pelagic fish of commercial importance. However, despite recent advances in multifrequency processing, the technique still requires some subjective judgement when allocating the acoustic data, fish-school ech...
39 CitationsSource
#1Alex De Robertis (NMFS: National Marine Fisheries Service)H-Index: 20
#2Ian Higginbottom (Hobart Corporation)H-Index: 3
A simple and effective post-processing technique to estimate echosounder background-noise levels and signal-to-noise ratios (SNRs) during active pinging is developed. Similar to other methods of noise estimation during active pinging, this method assumes that some portion of the sampled acoustic signal is dominated by background noise, with a negligible contribution from the backscattered transmit signal. If this assumption is met, the method will provide robust and accurate estimates of backgro...
87 CitationsSource
#1C. I. H. Anderson (UW: University of Washington)H-Index: 1
#2John K. Horne (UW: University of Washington)H-Index: 25
Last. J. BoyleH-Index: 1
view all 3 authors...
18 CitationsSource
An investigation of midwater zooplankton aggregations in a coastal fjord was conducted in November 2002. This study focused on quantitative comparisons between a calibrated, three-frequency (38, 120, and 200 kHz) vessel-based echo-sounder, a multinet towed zooplankton sampler (BIONESS), and a high-resolution underwater camera (ZOOVIS). Daytime layers of euphausiids and amphipods near 70–90-m depth were observed in lower parts of the inlet, especially concentrated by tidal flows around a sill. Qu...
34 CitationsSource
454 CitationsSource
#1David J. Miller (PSU: Pennsylvania State University)H-Index: 25
#2J. Browning (PSU: Pennsylvania State University)H-Index: 1
Several authors have shown that, when labeled data are scarce, improved classifiers can be built by augmenting the training set with a large set of unlabeled examples and then performing suitable learning. These works assume each unlabeled sample originates from one of the (known) classes. Here, we assume each unlabeled sample comes from either a known or from a heretofore undiscovered class. We propose a novel mixture model which treats as observed data not only the feature vector and the class...
79 CitationsSource
#1John K. Horne (UM: University of Michigan)H-Index: 25
152 CitationsSource
We propose an assessing method of mixture model in a cluster analysis setting with integrated completed likelihood. For this purpose, the observed data are assigned to unknown clusters using a maximum a posteriori operator. Then, the integrated completed likelihood (ICL) is approximated using the Bayesian information criterion (BIC). Numerical experiments on simulated and real data of the resulting ICL criterion show that it performs well both for choosing a mixture model and a relevant number o...
776 CitationsSource
Cited By14
Newest
#1Olav Brautaset (Norwegian Computing Center)H-Index: 1
#2Anders U. Waldeland (Norwegian Computing Center)H-Index: 3
Last. Nils Olav HandegardH-Index: 20
view all 7 authors...
1 CitationsSource
#1Alex De Robertis (NOAA: National Oceanic and Atmospheric Administration)H-Index: 20
#2Olav A. Ormseth (NOAA: National Oceanic and Atmospheric Administration)H-Index: 2
Abstract A series of daytime replicate (spring/summer/fall) acoustic surveys were conducted at 11 inshore sites along the Kodiak Island/Kenai Peninsula area and the outer coast of Southeast Alaska as part of the Gulf of Alaska Integrated Ecosystem Research Program in 2010, 2011, and 2013. A two-frequency technique was used to classify backscatter as ‘fish’ or’macrozooplankton’ based on the observed relative frequency response, which are used as proxies for the abundance of fish with swimbladders...
1 CitationsSource
#1Christopher Bassett (NOAA: National Oceanic and Atmospheric Administration)H-Index: 1
#2Alex De Robertis (NOAA: National Oceanic and Atmospheric Administration)H-Index: 20
Last. Christopher D. Wilson (NOAA: National Oceanic and Atmospheric Administration)H-Index: 16
view all 3 authors...
2 CitationsSource
Abstract Clustering is a useful unsupervised technique for the identification of acoustic groups in multi-frequency echograms based on frequency response. K-Means is the most well-known clustering technique but has significant requirements such as clusters of equal size and spherical shape. Initialization is a common problem in clustering as only local minima are usually guaranteed, and thus initialization must locate the centroids near the global minimum. Expectation-Maximization (EM) clusterin...
1 CitationsSource
#1Antonio Lopez-Serrano (IPN: Instituto Politécnico Nacional)H-Index: 1
#2Héctor Villalobos (IPN: Instituto Politécnico Nacional)H-Index: 4
Last. Manuel O. Nevárez-Martínez (Crips)H-Index: 13
view all 3 authors...
To obtain reliable fish biomass estimates by acoustic methods, it is essential to filter out the signals from unwanted scatterers (e.g. zooplankton). When acoustic data are collected at more than one frequency, methods that exploit the differences in reflectivity of scatterers can be used to achieve the separation of targets. These methods cannot be applied with historical data nor recent data collected on board fishing vessels employed as scientific platforms, where only one transducer is avail...
1 CitationsSource
#1Sven Gastauer (Curtin University)H-Index: 4
#2Ben Scoulding (Hobart Corporation)H-Index: 4
Last. Miles Parsons (Curtin University)H-Index: 8
view all 3 authors...
Ongoing monitoring of complex, mixed species environments is a challenging task. In this study, the potential of acoustic and catch data collected aboard a commercial fishing vessel, in combination with geostatistical variance estimates, are explored as a means to derive information on the distribution and abundance of key species groups within selected fishing regions. The FV Carolina M, a trap fishing vessel which operates in waters off Broome, Western Australia, in the Northern Demersal Scale...
1 CitationsSource
#1Matthew H. Pinkerton (NIWA: National Institute of Water and Atmospheric Research)H-Index: 15
5 CitationsSource
#1Sven Gastauer (Curtin University)H-Index: 4
#2Ben Scoulding (Hobart Corporation)H-Index: 4
Last. Miles Parsons (Curtin University)H-Index: 8
view all 3 authors...
Fisheries acoustics is now a standard tool for monitoring marine organisms. Another use of active-acoustics techniques is the potential to qualitatively describe fish school and seafloor characteristics or the distribution of fish density hotspots. Here, we use a geostatistical approach to describe the distribution of acoustic density hotspots within three fishing regions of the Northern Demersal Scalefish Fishery in Western Australia. This revealed a patchy distribution of hotspots within the t...
1 CitationsSource
#1Gary Vargas (UFRPE: Universidade Federal Rural de Pernambuco)H-Index: 2
#2Flávia Lucena-Frédou (UFRPE: Universidade Federal Rural de Pernambuco)H-Index: 7
Last. Arnaud BertrandH-Index: 2
view all 6 authors...
Underwater acoustics have an unrealized potential for multicomponent ecosystem characterization. A variety of methods (e.g., backscattering differences, clustering, Gaussian model) are currently used for multifrequency classification. In this study we propose and implement a new method based on the distribution of scatters on multifrequency spatial planes. Our approach is based on both the sum and the difference of backscattered energy of up to five bi-frequency pairs. This method was developed ...
Source
Source