Practical guidance on automated sorting of underwater images in plankton ecology research
Bibliographic record
Abstract
Abstract In situ plankton imaging complements classical sampling approaches by obtaining observations of plankton composition and traits at finer spatial and temporal resolutions. These imaging techniques can provide valuable ecological insight and are also notorious for generating a massive volume of images that require classification to generate quantitative data. Automating image segmentation and classification can accelerate data extraction; however, the high diversity and uneven distribution of plankton taxa, variation in image characteristics obtained from different imagers, and limited availability of human classification expertise present challenges to development of user-friendly and universally accepted image processing and classification tools. Differences in desired taxonomic or trait resolution, classifier performance, and spatial variability in community composition often necessitate the development of tailored automated classifiers for specific cases. This customization typically requires expertise in computer vision and machine learning that many ecologists do not acquire through traditional training. In this paper, we review the plankton imaging and classification workflow and present two case studies for a plankton ecology audience. We emphasize Convolutional Neural Network (CNN) classifiers and demonstrate strategies to address common challenges in image classification using semiautomated classification and unsupervised learning approaches. The overarching aim is to provide practical guidance for ecologists and encourage broader adoption of in situ plankton imaging in ecological research.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.011 | 0.028 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.004 | 0.004 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.003 | 0.002 |
| Insufficient payload (model declined to judge) | 0.017 | 0.021 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".