cs.CVFeb 3, 2026

Vision Transformers for Zero-Shot Clustering of Animal Images: A Comparative Benchmarking Study

Authors: Hugo Markoff, Stefan Hein Bengtson, Michael Ørsted

Organizations: Department of Chemistry and Bioscience, Aalborg University, Aalborg, Denmark · Department of Architecture, Design and Media Technology, Visual Analysis and Perception Lab, Aalborg University, Aalborg, Denmark

Abstract

Manual labelling of animal images remains a significant bottleneck in ecological research, limiting the scale and efficiency of biodiversity monitoring efforts. This study investigates whether state-of-the-art Vision Transformer (ViT) foundation models can reduce thousands of unlabelled animal images directly to species-level clusters. We present a comprehensive benchmarking framework evaluating five ViT models combined with five dimensionality reduction techniques and four clustering algorithms, two supervised and two unsupervised, across 60 species (30 mammals and 30 birds), with each test using a random subset of 200 validated images per species. We investigate when clustering succeeds at species-level, where it fails, and whether clustering within the species-level reveals ecologically meaningful patterns such as sex, age, or phenotypic variation. Our primary, fully unsupervised finding is that DINOv3 embeddings with t-SNE and HDBSCAN reach a species-level V-measure of 0.943 with no prior knowledge of the number of species, rejecting only 1.14% of images as outliers requiring expert review. Assigning one label per cluster reduces manual labelling actions by 98.3%, with 94.6% of clustered images labelled correctly, close to the accuracy of manual annotation. When the true number of species is supplied, supervised hierarchical clustering raises this to 0.958, which we report as an upper bound as the species count is typically unknown in real deployments. We further demonstrate robustness to realistic long-tailed distributions of species and show that intentional over-clustering can reveal intra-specific variation including age classes, sexual dimorphism, and pelage differences. We introduce an open-source benchmarking toolkit and provide recommendations for ecologists to select appropriate methods for sorting their specific taxonomic groups and data.

Explore similar work

Sep 15, 2026

InfoTaxa: Information-Calibrated Label-Free Clustering for Fine-Grained Visual Taxonomy

Label-free clustering of frozen pretrained visual embeddings offers a scalable route to biodiversity monitoring, but image-only fine-grained taxonomy exhibits a consistent coarse-to-fine failure mode: clusters recover broad taxonomic structure yet plateau at species level. We study this behaviour on BIOSCAN-5M through an information-calibrated clustering analysis. BioCLIP~2 features with UMAP and HDBSCAN reach 0.790.79 AMI at family and 0.670.67 at genus, substantially improving over the prior image baseline and remaining competitive with oracle-KK, graph-based, and learned clustering heads on the same frozen features. To diagnose whether the remaining plateau is method-limited or information-limited, we introduce InfoTaxa, which combines clustering efficiency---the fraction of probe-estimated image information recovered by an unsupervised partition---with paired DNA as an audit signal only, not an inference input. The density pipeline recovers approximately 0.900.90 and 0.810.81 of the image-available information at order and family, respectively. Held-out late-fusion probes show that adding DNA to the image embedding reduces species-level prediction error by approximately two bits. Robustness analyses cover multiple image encoders, described-species and rare-class subsets, probe diagnostics, and held-out-species coarse-rank generalisation and same-species retrieval. Thus, in the tested setting, species-level label-free clustering is both clustering-limited and representation-limited: improved clustering may recover additional image-exposed structure, but cannot close the DNA-audited information gap alone.
Sep 10, 2026

Can Edge-Deployable Vision-Language Models Identify Species?

Camera traps often run in the field on edge hardware with limited or no connectivity, making small, locally-deployable vision-language models (VLMs) -- not frontier-scale ones -- the practically relevant class to evaluate for species identification. We test whether models in this deployment-relevant 2--8B range carry genuine taxonomic knowledge, evaluating four such VLMs (Qwen3-VL 2B/4B/8B, Gemma3 4B) against the domain-specific specialist BioCLIP (300M parameters) on a 96-species task, comparing clean iNaturalist photographs against camera-trap imagery from 6 LILA.science collections, on two independently-sampled evaluation sets. All models identify species far above chance, but every model -- general-purpose or specialist -- degrades sharply on field imagery (domain gaps of 9.6--26.6 percentage points, consistent across taxonomic levels and both evaluation sets), indicating the degradation reflects general image legibility rather than fine-grained discrimination failure. BioCLIP substantially outperforms every VLM tested (by 33.2--59.2 percentage points across an expanded 200-image sample for every model) despite its far smaller size, suggesting the gap reflects specialized training data rather than model scale; yet BioCLIP's own domain gap (18.0 points) is statistically indistinguishable from the best VLM's (22.3 points), suggesting the clean-to-field degradation itself is a property of the image-quality shift rather than a general-purpose-model weakness. Under open-set prompting, 5.9--9.6% of responses are syntactically valid but taxonomically nonexistent species names; the relative fabrication-rate ranking across models replicates exactly across both evaluation sets, a more robust finding than any single point estimate.
Jun 28, 2026

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

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.