PlanktonLake-CEREEP - A Freshwater Plankton Image Dataset with Semi-Automated Label Cleaning

Abstract

Plankton plays a fundamental role in aquatic ecosystems, influencing biogeochemical cycles and serving as a key food source for many organisms. Recent high-throughput imaging technologies enable the rapid acquisition of large volumes of microscopic images, creating new opportunities for monitoring planktonic ecosystems. However, the manual processing and annotation of the vast amounts of data generated by these devices remain time-consuming tasks. In this context, machine learning-based classification models offer a promising solution. In this data paper, we introduce a new labeled freshwater plankton dataset comprising approximately 88,000 images distributed across 43 taxa. We also detail the methods we applied to ensure a high quality labeling, partly relying on an automatic detection of wrong labels. Finally, we present a baseline convolutional neural network (CNN), which achieves a classification accuracy of 93% on our dataset.

A promptable foundation model enables automated multi-task dataset construction and real-time pose estimation in mice
J. Cobos, S. Thirard, M. Belkaid +1
bioRxiv2 Oct 2026PreprintCC BYFull textZoology & animal behaviourbioRxiv
Fluctuating growth rate and spatial diffusion shape plankton diversity
Giorgio Vittorio Visco, Kobe Simoens, Emanuele Pigani +4
arXiv30 Sep 2026PreprintCC BY 4.0Full textEvolutionary biologyarXiv
DNA-based monitoring uncovers contrasting salinity-driven diversity patterns of bacterial and protist plankton in the Baltic Sea area
Krzysztof T. Jurdzinski, Meike AC Latz, Anders Torstensson +7
Communications Biology7 Oct 2026Peer-reviewedCC BY 4.0Full textEcologyCrossref
Direct effects of plankton size are stronger than abiotic factors in driving variation in fish life history traits
Maël Gernez, Didier D. Gascuel, Alaia Morell +1
Progress in Oceanography20 Oct 2026Peer-reviewedCC BY 4.0Full textEcologyHAL
Microbial plankton in southern Hudson Bay and James Bay in late summer
E. Collins, Z. Kazmiruk
bioRxiv1 Nov 2026PreprintCC BY-NCFull textMicrobiologybioRxiv

Cite

BibTeX RIS

BibTeX

@misc{dechaumet2026planktonlakecereep,
  title = {{PlanktonLake-CEREEP - A Freshwater Plankton Image Dataset with Semi-Automated Label Cleaning}},
  author = {Léo Déchaumet and Carine Puppo and Samy Blusseau and Beatriz Marcotegui and Etienne Decencière and Gérard Lacroix and David Carmignac and Jean-François Le Galliard},
  howpublished = {HAL},
  year = {2026},
  month = aug,
  doi = {10.64898/2026.09.24.753096},
  url = {https://cnrs.hal.science/hal-05761800v1},
  note = {CC BY-NC 4.0},
}

RIS

TY  - UNPB
TI  - PlanktonLake-CEREEP - A Freshwater Plankton Image Dataset with Semi-Automated Label Cleaning
AU  - Léo Déchaumet
AU  - Carine Puppo
AU  - Samy Blusseau
AU  - Beatriz Marcotegui
AU  - Etienne Decencière
AU  - Gérard Lacroix
AU  - David Carmignac
AU  - Jean-François Le Galliard
PY  - 2026
DA  - 2026/08/22
PB  - HAL
DO  - 10.64898/2026.09.24.753096
UR  - https://cnrs.hal.science/hal-05761800v1
AB  - Plankton plays a fundamental role in aquatic ecosystems, influencing biogeochemical cycles and serving as a key food source for many organisms. Recent high-throughput imaging technologies enable the rapid acquisition of large volumes of microscopic images, creating new opportunities for monitoring planktonic ecosystems. However, the manual processing and annotation of the vast amounts of data generated by these devices remain time-consuming tasks. In this context, machine learning-based classification models offer a promising solution. In this data paper, we introduce a new labeled freshwater plankton dataset comprising approximately 88,000 images distributed across 43 taxa. We also detail the methods we applied to ensure a high quality labeling, partly relying on an automatic detection of wrong labels. Finally, we present a baseline convolutional neural network (CNN), which achieves a classification accuracy of 93% on our dataset.
ER  -

Abstract

Related

Cite