# Explainable Deep Learning Framework for Reliable Species-Level Classification Within the Genera Desmodesmus and Tetradesmus

## TL;DR

This paper introduces a transparent AI framework that accurately classifies microalgae species using microscope images, achieving high performance and interpretability.

## Contribution

The study proposes an explainable deep learning framework for species-level classification of green algae with high accuracy and biological interpretability.

## Key findings

- ResNet152V2 achieved a macro F1-score of 0.975, the highest among evaluated models.
- Visualization techniques confirmed models focused on biologically relevant features like cell walls and surface structures.
- The framework demonstrates reliable classification even with limited datasets.

## Abstract

The accurate identification of microalgae remains challenging due to imaging variability, physiological changes, and environmental factors rather than inherent morphological similarity. This study presents an interpretable artificial intelligence framework designed to classify green algae species from microscope images with both high accuracy and transparency. Twelve deep learning models were systematically trained and compared using standardized image processing on three distinct species—Desmodesmus flavescens, Desmodesmus subspicatus and Tetradesmus dimorphus. Among the evaluated architectures, several convolutional neural networks achieved strong performance, with macro-level precision, recall, and F1-scores generally exceeding 0.90, and the best-performing model (ResNet152V2) reaching a macro F1-score of 0.975. Visualization analyses confirmed that the models based their decisions on true biological features, such as cell walls and surface structures, rather than irrelevant background elements. These findings demonstrate that artificial intelligence can achieve near-perfect recognition of microalgae even from limited datasets while maintaining interpretability. The proposed framework provides a reproducible and biologically meaningful tool for digital taxonomy.

Microalgae are an evolutionarily ancient and morphologically diverse group of photosynthetic eukaryotes, with taxonomic resolution complicated by environmentally driven phenotypic plasticity. This study merges deep learning and explainable artificial intelligence (XAI) to establish a transparent, reliable, and biologically meaningful framework for green microalgae (Chlorophyta) classification. Microscope images from three morphologically distinct algal species—Desmodesmus flavescens, Desmodesmus subspicatus, and Tetradesmus dimorphus representing the genera Desmodesmus and Tetradesmus within Chlorophyta—were analyzed using twelve convolutional neural networks, including EfficientNet-B0–B7, DenseNet201, NASNetLarge, Xception, and ResNet152V2. A curated dataset comprising 3624 microscopic images from three Chlorophyta species was used, split into training, validation, and test subsets. All models were trained using standardized preprocessing and data augmentation procedures, including grayscale conversion, CLAHE-based contrast enhancement, rotation, flipping, and brightness normalization. The model’s performance was assessed using accuracy and loss metrics on independent test datasets, while interpretability was evaluated through saliency maps and Gradient-weighted Class Activation Mapping (Grad-CAM) visualizations. ResNet152V2 achieved the highest overall performance among all evaluated architectures, outperforming EfficientNet variants, NASNetLarge, and Xception in terms of macro F1-score. Visualization analysis showed that both Grad-CAM and saliency mapping consistently highlighted biologically relevant regions—including cell walls, surface ornamentation, and colony structures—confirming that the models relied on taxonomically meaningful features rather than background artifacts. The findings indicate that the integration of deep learning and XAI can attain consistently high test accuracy for microalgal species, even with constrained datasets. This approach enables automated taxonomy and supports biodiversity monitoring, ecological assessment, biomass optimization, and biodiesel production by integrating interpretability with high predictive accuracy.