Melika Baghooee

Computer vision and AI are now widely used for automated insect classification, but their potential for estimating other traits, such as biomass, is not yet fully explored. Insect biomass is a key measure of ecosystem function, informing ecosystem services, food webs, and environmental change. It is also used to track population trends and estimate the contribution of insects to ecosystem carbon. Currently, wet bulk biomass is used as a standard measure for insect monitoring because it is fast and practical. However, bulk biomass conflates individual contributions, obscuring biologically meaningful variation. Obtaining individual-level biomass is challenging as each specimen must be dried and weighed, making large-scale measurements practically intractable. Here, we propose EntoScan, an open-source, low-cost and standardized imaging system based on a modified flatbed scanner and Biomass Estimation in Entomology (BEEomass), a novel computer vision model that estimates dry biomass of terrestrial arthropods from images alone. BEEomass uses a physically motivated, scale-invariant representation of body size and mass that allows it to generalize across taxa differing widely in morphology. We show that this image-based approach provides accurate dry biomass estimates (R² > 0.95) while considerably reducing both the manual effort and specimen destruction compared to direct weighing. We then illustrate the practical utility of EntoScan and BEEomass through two proof-of-concept case studies: tracking temperature-induced size variation in laboratory-reared Drosophila and monitoring seasonal biomass dynamics in wild-caught Pachygnatha degeeri. Because they are affordable, scalable, and easy to use, our two methods have the potential to serve as standardized tools for monitoring insect biomass at scale, and ultimately improve our ability to quantify and predict ecosystem functions and services, supporting more effective conservation and agricultural management.

Although visual features have been the cornerstone of insect recognition, morphological traits are often overlooked by modern “scale-invariant” deep learning-based methods. One such trait is absolute body size, which is commonly used by entomologists to describe species. We propose a novel approach that integrates explicit size information into computer vision models. Our method improves predictions of a Convolutional Neural Network (ResNet) by incorporating the likelihood of an insect belonging to each class based on its size and the size distribution of the training data. We compare the performance of the proposed size-aware approach against a standard ResNet and a feature-fused ResNet model, demonstrating that incorporating explicit dimensional traits, such as body size, can enhance classification accuracy across both balanced and imbalanced dataset scenarios. Our results also show that incorporating size reduces classification error, especially in low-data scenarios and for hierarchically coarse, taxonomic misclassifications (e.g., errors at the family level). Despite being readily available and computationally inexpensive, size has, to our knowledge, not been used in this context before. We show that explicitly including size is both computationally efficient and practical, as it requires no specialized sensors and can be derived from standard image segmentation. This makes our approach highly scalable for automated monitoring of insects with camera traps. By acknowledging the importance of size, our approach also contributes to more robust and interpretable ecological modeling. This work opens new avenues for biologically informed AI applications in entomology and beyond.