A trainable object finder, selector and identifier for pollen, spores and other things: A step towards automated pollen recognition in lake sediments

Research output: Journal contributionsJournal articlesResearchpeer-review

Standard

A trainable object finder, selector and identifier for pollen, spores and other things: A step towards automated pollen recognition in lake sediments. / Theuerkauf, Martin; Siradze, Nia; Gillert, Alexander.
In: Holocene, Vol. 34, No. 3, 01.03.2024, p. 297-305.

Research output: Journal contributionsJournal articlesResearchpeer-review

Harvard

APA

Vancouver

Theuerkauf M, Siradze N, Gillert A. A trainable object finder, selector and identifier for pollen, spores and other things: A step towards automated pollen recognition in lake sediments. Holocene. 2024 Mar 1;34(3):297-305. Epub 2023 Nov 29. doi: 10.1177/09596836231211876

Bibtex

@article{4c86581e391549388beb25b8dbc1e110,
title = "A trainable object finder, selector and identifier for pollen, spores and other things: A step towards automated pollen recognition in lake sediments",
abstract = "Pollen records are the most important proxy for reconstructing past terrestrial vegetation. While new approaches for improved quantitative interpretation of pollen data have been developed over the last decades, the availability of pollen records remains mostly limited because pollen samples are still analysed manually, which is a time-consuming task and requires extensive training. Here, we present an approach for automated recognition of pollen and spores from lake sediments using deep convolutional neural networks and machine learning. The approach includes two stages. The detector first locates pollen and spores in the sample matrix, and the classifier then classifies these objects. We have trained the approach on two pollen datasets from two lakes in north-eastern Germany. So far, our approach is able to automatically recognise 10 pollen types and Lycopodium spores with high accuracy. As soon as more training data are available, more pollen and spore types can be added. The preparation of training data, the training of the neural networks and their application are accessible via a freely available, browser-based user interface called TOFSI.",
keywords = "Automated pollen recognition, deep convolutional neural networks (DCNN), pollen analysis, Biology",
author = "Martin Theuerkauf and Nia Siradze and Alexander Gillert",
note = "Publisher Copyright: {\textcopyright} The Author(s) 2023.",
year = "2024",
month = mar,
day = "1",
doi = "10.1177/09596836231211876",
language = "English",
volume = "34",
pages = "297--305",
journal = "Holocene",
issn = "0959-6836",
publisher = "SAGE Publications Inc.",
number = "3",

}

RIS

TY - JOUR

T1 - A trainable object finder, selector and identifier for pollen, spores and other things

T2 - A step towards automated pollen recognition in lake sediments

AU - Theuerkauf, Martin

AU - Siradze, Nia

AU - Gillert, Alexander

N1 - Publisher Copyright: © The Author(s) 2023.

PY - 2024/3/1

Y1 - 2024/3/1

N2 - Pollen records are the most important proxy for reconstructing past terrestrial vegetation. While new approaches for improved quantitative interpretation of pollen data have been developed over the last decades, the availability of pollen records remains mostly limited because pollen samples are still analysed manually, which is a time-consuming task and requires extensive training. Here, we present an approach for automated recognition of pollen and spores from lake sediments using deep convolutional neural networks and machine learning. The approach includes two stages. The detector first locates pollen and spores in the sample matrix, and the classifier then classifies these objects. We have trained the approach on two pollen datasets from two lakes in north-eastern Germany. So far, our approach is able to automatically recognise 10 pollen types and Lycopodium spores with high accuracy. As soon as more training data are available, more pollen and spore types can be added. The preparation of training data, the training of the neural networks and their application are accessible via a freely available, browser-based user interface called TOFSI.

AB - Pollen records are the most important proxy for reconstructing past terrestrial vegetation. While new approaches for improved quantitative interpretation of pollen data have been developed over the last decades, the availability of pollen records remains mostly limited because pollen samples are still analysed manually, which is a time-consuming task and requires extensive training. Here, we present an approach for automated recognition of pollen and spores from lake sediments using deep convolutional neural networks and machine learning. The approach includes two stages. The detector first locates pollen and spores in the sample matrix, and the classifier then classifies these objects. We have trained the approach on two pollen datasets from two lakes in north-eastern Germany. So far, our approach is able to automatically recognise 10 pollen types and Lycopodium spores with high accuracy. As soon as more training data are available, more pollen and spore types can be added. The preparation of training data, the training of the neural networks and their application are accessible via a freely available, browser-based user interface called TOFSI.

KW - Automated pollen recognition

KW - deep convolutional neural networks (DCNN)

KW - pollen analysis

KW - Biology

UR - http://www.scopus.com/inward/record.url?scp=85178404431&partnerID=8YFLogxK

UR - https://www.mendeley.com/catalogue/4c2015c3-e8fe-3339-9ce5-f5f8268459f3/

U2 - 10.1177/09596836231211876

DO - 10.1177/09596836231211876

M3 - Journal articles

AN - SCOPUS:85178404431

VL - 34

SP - 297

EP - 305

JO - Holocene

JF - Holocene

SN - 0959-6836

IS - 3

ER -

Recently viewed

Researchers

  1. Nigel Forrest

Publications

  1. The Diffusion of Values among Democracies and Autocracies
  2. Belowground top-down and aboveground bottom-up effects structure multitrophic community relationships in a biodiverse forest
  3. Differential mortality rates in major and subthreshold depression
  4. You Are Where You Eat: A Theoretical Perspective on Why Identity Matters in Local Food Groups
  5. Cultural change in Asia and beyond
  6. Søren Kierkegaard in deutscher Sprache
  7. On entrepreneurial risk-taking and the macroeconomic effects of financial constraints
  8. Tree phylogenetic diversity promotes host–parasitoid interactions
  9. Big Data - Characterizing an Emerging Research Field using Topic Models
  10. Der Struwwelpeter
  11. When it really counts
  12. Exploring the Capacity of Water Framework Directive Indices to Assess Ecosystem Services in Fluvial and Riparian Systems
  13. Long-term retrospective analysis of the societal metabolism of cobalt in the European Union
  14. Introduction: Modeling the Pacific Ocean
  15. Lernsoftware im Unterricht
  16. Intra-specific leaf trait responses to species richness at two different local scales
  17. An Integrated Case Study of the Concepts and Applications of SAP ERP HCM
  18. Balanced scorecard and controllability at the level of middle managers
  19. From GUI to No-UI
  20. Export diversification and income differences reconsidered
  21. Knowledge on global environmental change within social praxis: what do we know?
  22. Using Reading Strategy Training to Foster Students´ Mathematical Modelling Competencies
  23. From Allegiant to Assertive Citizens
  24. Sol-gel technology for greener and more sustainable antimicrobial textiles that use silica matrices with C, and Ag and ZnO as biocides
  25. Insights into an Action-Oriented Training Program to Promote Sustainable Entrepreneurship
  26. The effectiveness of nudging
  27. Work in Progres Work on Progress
  28. Minimum return guarantees, investment caps, and investment flexibility
  29. Open Access und Open Educational Resources
  30. Linking socio-technical transition studies and organisational change management