<i>Zooglider:</i> An autonomous vehicle for optical and acoustic sensing of zooplankton

Ohman, Mark D.; Davis, Russ E; Sherman, Jeffrey T.; Grindley, Kyle; Whitmore, Benjamin M.; Nickels, Catherine F.; Ellen, Jeffrey

doi:10.1002/lom3.10301

Cited by 97 publications

(100 citation statements)

References 65 publications

(71 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The burgeoning number of digital imaging methods available to aquatic ecologists, both in situ (Davis et al ; Samson et al ; Benfield et al ; Watson ; Olson and Sosik ; Cowen and Guigland ; Picheral et al ; Schulz et al ; Thompson et al ; Briseño‐Avena et al ; Ohman et al ) and in the laboratory (Sieracki et al ; Gorsky et al ), is generating rapidly expanding libraries of digital images useful in a variety of scientific applications. However, the accumulation of large numbers of images increases the need for much more efficient machine learning methods in order to automate the processes of image classification, data extraction, and analysis.…”

mentioning

confidence: 99%

Improving plankton image classification using context metadata

Ellen

Graff

Ohman

2019

Limnology & Ocean Methods

Self Cite

View full text Add to dashboard Cite

Advances in both hardware and software are enabling rapid proliferation of in situ plankton imaging methods, requiring more effective machine learning approaches to image classification. Deep Learning methods, such as convolutional neural networks (CNNs), show marked improvement over traditional feature-based supervised machine learning algorithms, but require careful optimization of hyperparameters and adequate training sets. Here, we document some best practices in applying CNNs to zooplankton and marine snow images and note where our results differ from contemporary Deep Learning findings in other domains. We boost the performance of CNN classifiers by incorporating metadata of different types and illustrate how to assimilate metadata beyond simple concatenation. We utilize both geotemporal (e.g., sample depth, location, time of day) and hydrographic (e.g., temperature, salinity, chlorophyll a) metadata and show that either type by itself, or both combined, can substantially reduce error rates. Incorporation of context metadata also boosts performance of the feature-based classifiers we evaluated: Random Forest, Extremely Randomized Trees, Gradient Boosted Classifier, Support Vector Machines, and Multilayer Perceptron. For our assessments, we use an original data set of 350,000 in situ images (roughly 50% marine snow and 50% nonsnow sorted into 26 categories) from a novel in situ Zooglider. We document asymptotically increasing performance with more computationally intensive techniques, such as substantially deeper networks and artificially augmented data sets. Our best model achieves 92.3% accuracy with our 27-class data set. We provide guidance for further refinements that are likely to provide additional gains in classifier accuracy.

show abstract

mentioning

confidence: 99%

Improving plankton image classification using context metadata

Ellen

Graff

Ohman

2019

Limnology & Ocean Methods

Self Cite

View full text Add to dashboard Cite

show abstract

“…Recently, new sensors, instruments, platforms (e.g., Argo floats see Roemmich et al, 2019, and gliders see Testor et al, 2019) and methods (imaging, acoustics, omics, etc.) have been developed and deployed to improve the spatiotemporal resolution of planktonic communities and particles (Powell and Ohman, 2015a;Brownlee et al, 2016;Hunter-Cevera et al, 2016;Ohman et al, 2018) even in hostile conditions (Grossmann et al, 2015). Here we suggest that with these technologies (as well as concurrent advances in software), we have the ability to collect and analyze significantly more global information on plankton distributions and diversity and at a finer scale (taxonomic, spatial, and temporal) than is currently done.…”

Section: Context and Rationale Why Should We Observe Plankton And Parmentioning

confidence: 99%

Globally Consistent Quantitative Observations of Planktonic Ecosystems

et al. 2019

Self Cite

View full text Add to dashboard Cite

In this paper we review the technologies available to make globally quantitative observations of particles in general-and plankton in particular-in the world oceans, and for sizes varying from sub-microns to centimeters. Some of these technologies have been available for years while others have only recently emerged. Use of these technologies is critical to improve understanding of the processes that control abundances, distributions and composition of plankton, provide data necessary to constrain and improve ecosystem and biogeochemical models, and forecast changes in marine ecosystems in light of climate change. In this paper we begin by providing the motivation for plankton observations, quantification and diversity qualification on a global scale. We then expand on the state-of-the-art, detailing a variety of relevant and (mostly) mature technologies and measurements, including bulk measurements of plankton, pigment composition, uses of genomic, optical and acoustical methods as well

show abstract

“…Most current plankton samplers are far too fragile for Ships of Opportunity (SOOP) and require dedicated research ship time, making them far too expensive for long-term, large-scale surveys. While autonomous samplers that can cover reasonably large distances are in the pilot phase (e.g., Ohman et al, 2019) there are still significant start-up, maintenance, and data processing costs. The expense of microscopy required to process CPR samples is offset by the longevity of the instrument.…”

Section: New Surveys To Fill Gapsmentioning

confidence: 99%