Cross Granularity Metric Fusion for Rare Species Identification Under Scarce Biodiversity Image Data Tasks
Keywords:
Fine-Grained Visual Categorization, Deep Metric Learning, Rare Species, Feature Fusion, Biodiversity InformaticsAbstract
Biodiversity conservation increasingly relies on automated image identification systems to monitor ecosystems and track the population dynamics of vulnerable taxa. However, rare species identification presents profound computational challenges due to extreme data scarcity, pronounced intraclass variance caused by environmental factors, and subtle interclass visual differences among closely related species. Standard convolutional neural networks and traditional metric learning approaches frequently fail to capture the multi-level discriminative features necessary for accurate taxonomic classification under such constrained conditions. This paper introduces a novel cross-granularity metric fusion framework designed specifically for the visual identification of rare and endangered species. By extracting and aligning semantic features across multiple spatial resolutions and hierarchical levels, the proposed methodology constructs a highly robust feature representation that mitigates the adverse effects of long-tailed data distributions inherent in ecological datasets. We detail a comprehensive architectural design that simultaneously processes macroscopic morphological traits alongside microscopic texture patterns, subsequently fusing these distinct granularities through a unified distance metric space. Extensive evaluations on established fine-grained ecological datasets demonstrate that our approach significantly outperforms contemporary baseline architectures in top-1 accuracy and mean average precision, particularly for severely underrepresented taxonomic classes. Furthermore, comprehensive ablation studies validate the critical necessity of hierarchical feature integration and the effectiveness of the proposed hybrid distance optimization strategy. The findings suggest that bridging spatial and semantic granularities in deep metric learning offers a highly viable and scalable pathway for automated ecological monitoring applications where data acquisition is fundamentally limited.
References
1. Xu, W., Huang, H., Gong, Y., Yu, L., Wu, Q., & Zhang, J. (2025). Hierarchical multi-prototype discrimination: Boosting support-query matching for few-shot segmentation. IEEE Transactions on Multimedia.
2. Guo, C., Xiao, Z., Liu, Z., & Peng, S. (2020). Rg conditional diagnosability: a novel generalized measure of system-level diagnosis. Theoretical Computer Science, 814, 19-27.
3. Zhou, D. (2026). AI-Driven Hybrid SAST–DAST–SCA–IAST Framework for Risk-Based Vulnerability Prioritization in Microservice Architectures.
4. Tu, P., Huang, Y., Zheng, F., He, Z., Cao, L., & Shao, L. (2022, June). Guidedmix-net: Semi-supervised semantic segmentation by using labeled images as reference. In Proceedings of the AAAI conference on artificial intelligence (Vol. 36, No. 2, pp. 2379-2387).
5. Cong, P., Xu, Y., Ren, Y., Zhang, J., Xu, L., Wang, J., ... & Ma, Y. (2023, June). Weakly supervised 3d multi-person pose estimation for large-scale scenes based on monocular camera and single lidar. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 37, No. 1, pp. 461-469).
6. Song, S., Tang, Y., & Qin, R. (2025). Synthetic Data Matters: Re-training with Geo-typical Synthetic Labels for Building Detection. IEEE Transactions on Geoscience and Remote Sensing.
7. Wang, C., Ke, L., Liu, X., Huang, D., Yao, J., & Chu, W. Explainable AI Analysis of Brake Control in CARLA Through Reference and Distilled Policies. Frontiers in Robotics and AI, 13, 1847808.
8. Ding, Y., Li, Z., Huang, D., Zhang, K., Li, Z., & Feng, W. (2022). Adaptive range guided multi-view depth estimation with normal ranking loss. In Proceedings of the Asian Conference on Computer Vision (pp. 1892-1908).
9. Guo, Zixin, Kai Zhao, and Luyan Zhang. "InstanceRSR: Real-World Super-Resolution via Instance-Aware Representation Alignment." ICASSP 2026-2026 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2026.
10. Ma, Y., Qu, D., & Pyrozhenko, M. (2026). Bio-RegNet: A Meta-Homeostatic Bayesian Neural Network Framework Integrating Treg-Inspired Immunoregulation and Autophagic Optimization for Adaptive Community Detection and Stable Intelligence. Biomimetics, 11(1), 48.
11. Li, Z., Bai, X., Lu, J., Shen, P., Lam, E. Y., & Peng, Y. (2026). EventTracer: Fast Path Tracing-based Event Stream Rendering. IEEE Transactions on Visualization and Computer Graphics.
12. Li, Q., Luo, T., & Liao, J. (2025). Theory-inspired deep multi-view multi-label learning with incomplete views and noisy labels. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 20706-20715).
13. Qi, Z., Wang, S., Zhang, W., & Huang, Q. (2024). Uncertainty-boosted robust video activity anticipation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(12), 7775-7792.
14. Chen, L., Jiang, X., Liu, X., & Haardt, M. (2022). Reweighted low-rank factorization with deep prior for image restoration. IEEE Transactions on Signal Processing, 70, 3514-3529.
15. Peng, Q., Zheng, C., & Chen, C. (2023). Source-free domain adaptive human pose estimation. In Proceedings of the IEEE/CVF International Conference on Computer Vision (pp. 4826-4836).
16. Peng, Q., Xue, H., Wang, P., & Chen, C. (2026, March). Lifelong Domain Adaptive 3D Human Pose Estimation. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 40, No. 10, pp. 8358-8366).
17. Chu, Y. B., Li, J., Jia, P., Cui, J., Zhang, R., Kang, X., Lv, M., & Zhang, S. (2021). Irf1- and Egr1-activated transcription plays a key role in macrophage polarization: A multiomics sequencing study with partial validation. International immunopharmacology, 99, 108072. https://doi.org/10.1016/j.intimp.2021.108072
18. Ziyue Wang, Sheng Jin, Zhongrong Zuo, Jiawei Wu, Han Qiu, Qi She, Hao Zhang, and Xudong Jiang. Video-ktr: Re?inforcing video reasoning via key token attribution. arXiv preprint arXiv:2601.19686, 2026.
19. Deng, L., Deng, S., Chen, Y., Dai, Y., Zhong, Z., Li, L., ... & Huang, H. (2026). Tool-Augmented Agent for Closed-loop Optimization, Simulation, and Modeling Orchestration. arXiv preprint arXiv:2605.20190.
20. Li, Q., Liu, Z., Luo, W., Luo, T., & Hou, C. (2026). Correcting Visual Blur Induced by Attention Distraction to Reduce Hallucinations: Algorithm and Theory. arXiv preprint arXiv:2605.24602.
21. Qu, W., Shao, Y., Meng, L., Huang, X., & Xiao, L. (2024). A conditional denoising diffusion probabilistic model for point cloud upsampling. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 20786-20795).
22. Liu, Y., Liu, H., Wang, H., Meng, F., & Liu, M. (2023). BCAN: Bidirectional correct attention network for cross-modal retrieval. IEEE Transactions on Neural Networks and Learning Systems, 35(10), 14247-14258.
23. Li, Z., Chen, Z., Qu, F., Wang, M., Zhao, Y., Zhang, K., & Peng, Y. (2024). Point resampling and ray transformation aid to editable NeRF models. arXiv preprint arXiv:2405.07306.
24. Zhao, Haoyu, et al. "LSTD: Long Short-Term Temporal Diffusion for Video Generation." IEEE Transactions on Multimedia (2026).
25. Yang, Y., Ma, X., Li, C., Zheng, Z., Zhang, Q., Huang, G., ... & Zhao, Q. (2021). Believe what you see: Implicit constraint approach for offline multi-agent reinforcement learning. Advances in Neural Information Processing Systems, 34, 10299-10312.
26. Lv, Q., Kong, W., Li, H., Zeng, J., Qiu, Z., Qu, D., ... & Pang, J. (2025). F1: A vision-language-action model bridging understanding and generation to actions. arXiv preprint arXiv:2509.06951.
27. Qu, W., Wang, J., Gong, Y., Huang, X., & Xiao, L. (2025). An end-to-end robust point cloud semantic segmentation network with single-step conditional diffusion models. In Proceedings of the Computer Vision and Pattern Recognition Conference (pp. 27325-27335).
28. Agrawal, H., Huang, X., Kim, J. H., Li, Y., Ma, Y., Ning, T., & Tao, Y. (2022). Automated image processing and insight presentation. Patent / invention record. Publication date: 2022-04-05.
29. Huang, H., Zhang, J., Zhang, J., Xu, J., & Wu, Q. (2020). Low-rank pairwise alignment bilinear network for few-shot fine-grained image classification. IEEE Transactions on Multimedia, 23, 1666-1680.
30. Tang, A., Hiraoka, T., Hiraoka, N., Shi, F., Kawaharazuka, K., Kojima, K., ... & Inaba, M. (2024, May). Humanmimic: Learning natural locomotion and transitions for humanoid robot via wasserstein adversarial imitation. In 2024 IEEE International Conference on Robotics and Automation (ICRA) (pp. 13107-13114). IEEE.
31. Sang, Y. (2025, July). Towards explainable rag: Interpreting the influence of retrieved passages on generation. In 2025 4th International Conference on Robotics, Artificial Intelligence and Intelligent Control (RAIIC) (pp. 397-400). IEEE.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Moritz Keller, Violet Bennett, Roman Lebedev
This work is licensed under a Creative Commons Attribution 4.0 International License.
