How many labels do you need? A decision framework for cross-habitat marine species recognition

2026-06-27

AI Summary

Deploying species identification systems across diverse marine habitats faces high annotation costs and a lack of empirical guidance. This study proposes a decision framework to systematically evaluate the transfer performance of models—including DINOv2, CLIP, ResNet-50, and EfficientNet-B4—across multiple locations and taxonomic groups, using strategies such as linear probing, LoRA, visual prompt tuning, and full fine-tuning. The findings demonstrate that freezing the DINOv2 backbone and attaching a lightweight linear classifier achieves robust cross-habitat generalization with only 10–20 labeled images per species. This approach avoids learning habitat-specific shortcuts, reduces annotation requirements by an order of magnitude, and matches the performance of models trained with extensive fine-tuning.

Abstract

Automated image recognition is increasingly used to scale ecological monitoring beyond manual annotation, yet ecologists lack evidence-based guidance on how much labelling effort reliable deployment at new sites requires. We present a decision framework quantifying the trade-off between labelling effort and recognition accuracy when transferring vision systems across marine habitats. The benchmark spans five datasets, three oceans, and three taxonomic groups (fish, corals, invertebrates), from tropical reefs in the Great Barrier Reef and French Polynesia to a temperate Danish fjord. We evaluated four recognition models (DINOv2, CLIP, ResNet-50, EfficientNet-B4) under four adaptation strategies (linear probing, LoRA, Visual Prompt Tuning, full fine-tuning) across three protocols: within-habitat transfer across 20 reef sites (240 runs), cross-dataset geographic transfer along a difficulty gradient (40 runs), and few-shot adaptation curves with 0-100 labelled samples per class (648 runs). Frozen self-supervised foundation features (DINOv2 + linear classifier, 1,538 trainable parameters) generalised to unseen reef sites at least as well as fully fine-tuned convolutional baselines four orders of magnitude larger; they learned species-diagnostic, habitat-invariant representations, whereas baselines encoded habitat-specific shortcuts that fail at new sites. As few as 10-20 labelled images per species sufficed to deploy reliable recognition at a new site, cutting annotation effort by roughly an order of magnitude. Solution. Programmes expanding to new sites can deploy reliable recognition by pairing a frozen, open foundation model (DINOv2) with a simple linear classifier and annotating only 10-20 images per species - roughly 1-4 hours per site. The framework lets programmes budget labelling effort against expected accuracy across sites, ecosystems, and platforms.

Problem

• cross-habitat transfer
• marine species recognition
• labeling effort
• image recognition
• ecological monitoring

Innovation

• foundation models
• cross-habitat transfer
• few-shot adaptation
• self-supervised learning
• marine species recognition