Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.09690 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917400008458240 |
|---|---|
| author | Rueda-Toicen, Antonio Martin, Abigail Allen Morozov, Daniil Mahmood, Matin Schild, Alexandra Dayani, Shahabeddin Panza, Davide de Melo, Gerard |
| author_facet | Rueda-Toicen, Antonio Martin, Abigail Allen Morozov, Daniil Mahmood, Matin Schild, Alexandra Dayani, Shahabeddin Panza, Davide de Melo, Gerard |
| contents | Jaguar re-identification (re-ID) from citizen-science imagery can look strong on standard retrieval metrics while still relying on the wrong evidence, such as background context or silhouette shape, instead of the coat pattern that defines identity. We introduce a diagnostic framework for wildlife re-ID with two axes: a leakage-controlled context ratio, background/foreground, computed from inpainted background-only versus foreground-only images, and a laterality diagnostic based on cross-flank retrieval and mirror self-similarity. To make these diagnostics measurable, we curate a Pantanal jaguar benchmark with per-pixel segmentation masks and an identity-balanced evaluation protocol. We then use representative mitigation families, ArcFace fine-tuning, anti-symmetry regularization, and Lorentz hyperbolic embeddings, as case studies under the same evaluation lens. The goal is not only to ask which model ranks best, but also what visual evidence it uses to do so. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_09690 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Are We Recognizing the Jaguar or Its Background? A Diagnostic Framework for Jaguar Re-Identification Rueda-Toicen, Antonio Martin, Abigail Allen Morozov, Daniil Mahmood, Matin Schild, Alexandra Dayani, Shahabeddin Panza, Davide de Melo, Gerard Computer Vision and Pattern Recognition Jaguar re-identification (re-ID) from citizen-science imagery can look strong on standard retrieval metrics while still relying on the wrong evidence, such as background context or silhouette shape, instead of the coat pattern that defines identity. We introduce a diagnostic framework for wildlife re-ID with two axes: a leakage-controlled context ratio, background/foreground, computed from inpainted background-only versus foreground-only images, and a laterality diagnostic based on cross-flank retrieval and mirror self-similarity. To make these diagnostics measurable, we curate a Pantanal jaguar benchmark with per-pixel segmentation masks and an identity-balanced evaluation protocol. We then use representative mitigation families, ArcFace fine-tuning, anti-symmetry regularization, and Lorentz hyperbolic embeddings, as case studies under the same evaluation lens. The goal is not only to ask which model ranks best, but also what visual evidence it uses to do so. |
| title | Are We Recognizing the Jaguar or Its Background? A Diagnostic Framework for Jaguar Re-Identification |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.09690 |