On the impact of Citizen Science-derived data quality on deep learning based classification in marine images

Langenkämper D, Simon-Lledó E, Hosking B, Jones DOB, Nattkemper TW (2019)
PLOS ONE 14(6): e0218086.

Zeitschriftenaufsatz | Veröffentlicht | Englisch
 
Download
OA 3.28 MB
Autor*in
Langenkämper, DanielUniBi ; Simon-Lledó, Erik; Hosking, Brett; Jones, Daniel O. B.; Nattkemper, Tim WilhelmUniBi
Abstract / Bemerkung
The evaluation of large amounts of digital image data is of growing importance for biology, including for the exploration and monitoring of marine habitats. However, only a tiny percentage of the image data collected is evaluated by marine biologists who manually interpret and annotate the image contents, which can be slow and laborious. In order to overcome the bottleneck in image annotation, two strategies are increasingly proposed: “citizen science” and “machine learning”. In this study, we investigated how the combination of citizen science, to detect objects, and machine learning, to classify megafauna, could be used to automate annotation of underwater images. For this purpose, multiple large data sets of citizen science annotations with different degrees of common errors and inaccuracies observed in citizen science data were simulated by modifying “gold standard” annotations done by an experienced marine biologist. The parameters of the simulation were determined on the basis of two citizen science experiments. It allowed us to analyze the relationship between the outcome of a citizen science study and the quality of the classifications of a deep learning megafauna classifier. The results show great potential for combining citizen science with machine learning, provided that the participants are informed precisely about the annotation protocol. Inaccuracies in the position of the annotation had the most substantial influence on the classification accuracy, whereas the size of the marking and false positive detections had a smaller influence.
Erscheinungsjahr
2019
Zeitschriftentitel
PLOS ONE
Band
14
Ausgabe
6
Art.-Nr.
e0218086
ISSN
1932-6203
eISSN
1932-6203
Page URI
https://pub.uni-bielefeld.de/record/2936069

Zitieren

Langenkämper D, Simon-Lledó E, Hosking B, Jones DOB, Nattkemper TW. On the impact of Citizen Science-derived data quality on deep learning based classification in marine images. PLOS ONE. 2019;14(6): e0218086.
Langenkämper, D., Simon-Lledó, E., Hosking, B., Jones, D. O. B., & Nattkemper, T. W. (2019). On the impact of Citizen Science-derived data quality on deep learning based classification in marine images. PLOS ONE, 14(6), e0218086. doi:10.1371/journal.pone.0218086
Langenkämper, D., Simon-Lledó, E., Hosking, B., Jones, D. O. B., and Nattkemper, T. W. (2019). On the impact of Citizen Science-derived data quality on deep learning based classification in marine images. PLOS ONE 14:e0218086.
Langenkämper, D., et al., 2019. On the impact of Citizen Science-derived data quality on deep learning based classification in marine images. PLOS ONE, 14(6): e0218086.
D. Langenkämper, et al., “On the impact of Citizen Science-derived data quality on deep learning based classification in marine images”, PLOS ONE, vol. 14, 2019, : e0218086.
Langenkämper, D., Simon-Lledó, E., Hosking, B., Jones, D.O.B., Nattkemper, T.W.: On the impact of Citizen Science-derived data quality on deep learning based classification in marine images. PLOS ONE. 14, : e0218086 (2019).
Langenkämper, Daniel, Simon-Lledó, Erik, Hosking, Brett, Jones, Daniel O. B., and Nattkemper, Tim Wilhelm. “On the impact of Citizen Science-derived data quality on deep learning based classification in marine images”. PLOS ONE 14.6 (2019): e0218086.
Alle Dateien verfügbar unter der/den folgenden Lizenz(en):
Creative Commons Namensnennung 4.0 International Public License (CC-BY 4.0):
Volltext(e)
Access Level
OA Open Access
Zuletzt Hochgeladen
2019-07-17T08:25:27Z
MD5 Prüfsumme
3dbd5bde6e5a48fe0b47580e53a74358

8 References

Daten bereitgestellt von Europe PubMed Central.

Deep learning.
LeCun Y, Bengio Y, Hinton G., Nature 521(7553), 2015
PMID: 26017442
Biological responses to disturbance from simulated deep-sea polymetallic nodule mining.
Jones DO, Kaiser S, Sweetman AK, Smith CR, Menot L, Vink A, Trueblood D, Greinert J, Billett DS, Arbizu PM, Radziejewska T, Singh R, Ingole B, Stratmann T, Simon-Lledo E, Durden JM, Clark MR., PLoS ONE 12(2), 2017
PMID: 28178346
Semi-automated image analysis for the assessment of megafaunal densities at the Arctic deep-sea observatory HAUSGARTEN.
Schoening T, Bergmann M, Ontrup J, Taylor J, Dannheim J, Gutt J, Purser A, Nattkemper TW., PLoS ONE 7(6), 2012
PMID: 22719868
MAIA-A machine learning assisted image annotation method for environmental monitoring and exploration.
Zurowietz M, Langenkamper D, Hosking B, Ruhl HA, Nattkemper TW., PLoS ONE 13(11), 2018
PMID: 30444917
What Is Citizen Science?--A Scientometric Meta-Analysis.
Kullenberg C, Kasperowski D., PLoS ONE 11(1), 2016
PMID: 26766577
Deep learning is combined with massive-scale citizen science to improve large-scale image classification.
Sullivan DP, Winsnes CF, Akesson L, Hjelmare M, Wiking M, Schutten R, Campbell L, Leifsson H, Rhodes S, Nordgren A, Smith K, Revaz B, Finnbogason B, Szantner A, Lundberg E., Nat. Biotechnol. 36(9), 2018
PMID: 30125267
From principles to practice: a spatial approach to systematic conservation planning in the deep sea.
Wedding LM, Friedlander AM, Kittinger JN, Watling L, Gaines SD, Bennett M, Hardy SM, Smith CR., Proc. Biol. Sci. 280(1773), 2013
PMID: 24197407

Export

Markieren/ Markierung löschen
Markierte Publikationen

Open Data PUB

Web of Science

Dieser Datensatz im Web of Science®

Quellen

PMID: 31188894
PubMed | Europe PMC

Suchen in

Google Scholar