What question did this study set out to answer?

To develop and evaluate convolutional neural network pipelines for automatically identifying living and dead benthic foraminifera at the species level using Rose Bengal staining.

August 20, 2026Journal of MicropalaeontologyOpen Access

Identification of living (Rose Bengal)-stained benthic foraminifera using automated image recognition

View Full Paper

Ask AI

Bookmark

Authors

TWTobias Walla CBChristine Barras EGEmmanuelle Geslin

Discussion

Member takes

Overview

Validation study demonstrates accurate species identification and vital status classification in benthic foraminifera via neural networks, highlighting potential for rapid environmental biomonitoring.

Key Points

To develop and evaluate convolutional neural network pipelines for automatically identifying living and dead benthic foraminifera at the species level using Rose Bengal staining.
Collected low-diversity assemblages from the French Atlantic coast (Bourgneuf Bay) and high-diversity assemblages from the French Mediterranean coast.
Acquired specimen images using an automated setup comprising a modified 3D printer equipped with a camera.
Trained three convolutional neural network (CNN) models to classify species and distinguish vital status (living vs. dead) based on Rose Bengal staining.
The CNN models classified total community species with an accuracy of 96.0% in low-diversity settings and 82.2% in high-diversity settings.
A specialized CNN model distinguished both species identity and vital status in the low-diversity assemblage with 94.2% accuracy.
Model outputs aligned with published ecological quality indices used for biomonitoring high-diversity marine ecosystems.

Cite This Study

Walla et al. (2026) studied this question.

synapsesocial.com/papers/6a86b4cb8a91293e6a1cc4da https://doi.org/10.5194/jm-45-623-2026

View Full Paper

Ask AI

Bookmark

August 20, 2026Journal of MicropalaeontologyOpen Access

Identification of living (Rose Bengal)-stained benthic foraminifera using automated image recognition

View Full Paper

Ask AI

Bookmark

Authors

TWTobias Walla CBChristine Barras EGEmmanuelle Geslin

Discussion

Member takes

Overview

Key Points

To develop and evaluate convolutional neural network pipelines for automatically identifying living and dead benthic foraminifera at the species level using Rose Bengal staining.
Collected low-diversity assemblages from the French Atlantic coast (Bourgneuf Bay) and high-diversity assemblages from the French Mediterranean coast.
Acquired specimen images using an automated setup comprising a modified 3D printer equipped with a camera.
Trained three convolutional neural network (CNN) models to classify species and distinguish vital status (living vs. dead) based on Rose Bengal staining.
The CNN models classified total community species with an accuracy of 96.0% in low-diversity settings and 82.2% in high-diversity settings.
A specialized CNN model distinguished both species identity and vital status in the low-diversity assemblage with 94.2% accuracy.
Model outputs aligned with published ecological quality indices used for biomonitoring high-diversity marine ecosystems.

Cite This Study

Walla et al. (2026) studied this question.

synapsesocial.com/papers/6a86b4cb8a91293e6a1cc4da https://doi.org/10.5194/jm-45-623-2026

View Full Paper

Ask AI

Bookmark