Class-conditioned modeling improves recognition of unknown insect species
CUBE: Class-conditioned Unknown Boundary Evidence for Open-Set Insect Species Recognition
Computer Vision and Pattern Recognition
Summary
Identifying insect species can be tricky when some species look very similar and new species appear. The authors propose a method called CUBE that checks how likely an insect is to belong to a known species by comparing it locally against similar species rather than treating all unknowns as one group. This way, it better spots insects that are unknown but look close to known types. If additional data like location is available, their method CUBE-E uses it to improve accuracy further.
What this means in practice
- •For biodiversity monitoring teams: Improve automated insect species identification by detecting unknown species that are visually similar to known species, reducing misclassifications in ecological surveys.
- •For environmental consulting firms: Use class-conditioned evidence to validate insect species identifications in environmental impact assessments, especially when encountering unfamiliar species.
Authors
Hao Tan, Qimin Zhang, Zhe Xiao, Zekai Zuo, Wanting Liu, Jinghui Qin
Abstract
Open-set insect species recognition faces significant challenges when deployed in real-world biodiversity scenarios, particularly when encountering unknown species that are visually close to known training classes. Unlike generic out-of-distribution samples, fine-grained unknown insects are often not visually anomalous. They instead tend to appear near the decision boundary of specific known species, leading to unreliable closed-set predictions. To address these challenges, we propose \textbf{Class-conditioned Unknown Boundary Evidence (CUBE)}, a novel evidence-based framework that models unknownness around each known species rather than as a single global unknown class. Our key idea is to evaluate a test sample relative to the known species it most closely resembles, enabling local rejection for visually confusable unknown insects. CUBE learns class-conditioned boundary evidence from auxiliary species, preserves auxiliary species as shadow references during inference, and incorporates taxonomic candidate calibration to verify biological plausibility. When ecological metadata is available, we further introduce CUBE-E, which adds a lightweight location-consistency term without modifying the visual backbone. Experiments on the Open-Insect C-America benchmark show that CUBE achieves superior performance, outperforming state-of-the-art baselines with 95.8\%, 98.7\%, and 99.1\% AUROC on local, non-local, and non-moth unknown splits, respectively, while maintaining 94.9\% closed-set accuracy. With ecological metadata, CUBE-E further improves the results to 96.0\%, 99.7\%, and 99.5\% AUROC with 95.2\% closed-set accuracy.