Use this link to cite:
https://hdl.handle.net/2183/50150 Multi-View Analysis for Taxonomic Differentiation of Pimelia Beetle Species Using Deep Learning-Based Multi-Class Segmentation and Elytral Features
Loading...
Identifiers
Publication date
Authors
Ramallal, David
Mas-Peinado, Paloma
García-París, Mario
Advisors
Other responsabilities
Journal Title
Bibliographic citation
D. Ramallal, P. Mas-Peinado, M. García-París, M. Vila, J. De Moura and J. Novo, "Multi-View Analysis for Taxonomic Differentiation of Pimelia Beetle Species Using Deep Learning-Based Multi-Class Segmentation and Elytral Features," 2026 International Joint Conference on Neural Networks (IJCNN), Maastricht, Netherlands, 2026, pp. 1-8, doi: 10.1109/IJCNN69854.2026.11729995
Type of academic work
Academic degree
Abstract
[Abstract] The morphological differentiation of closely related beetle species in the genus Pimelia poses significant challenges for traditional taxonomic methods, which are labor-intensive and prone to subjectivity. This study introduces a novel framework that integrates multi-class segmentation, multi-view feature extraction, and taxonomic separability analysis to automate and enhance species classification. Unlike prior approaches, which primarily focus on direct classification tasks, our methodology emphasizes the quantification and visualization of interspecies separability. The proposed system employs deep learning architectures for precise multi-class segmentation of the elytral striae, followed by the extraction of robust multi-view features capturing texture, morphology, and structural variability. These features enable the computation of interspecies distances, visualized as heatmaps, providing a deeper understanding of taxonomic distinctions. Experimental results demonstrate the efficacy of this approach in accurately differentiating closely related species while offering an interpretable analysis of their morphological traits. This scalable and robust framework establishes a strong foundation for future advancements in biodiversity research, species conservation, and automated taxonomic classification.
Description
Presented at: 2026 International Joint Conference on Neural Networks (IJCNN), 22-26 June 2026, Maastricht, Netherlands
© 2026 IEEE. This version of the article has been accepted for publication, after peer review. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. The Version of Record is available online at: https://doi.org/10.1109/IJCNN69854.2026.11729995
Editor version
Rights
Copyright © 2026, IEEE







