On Thin Ice: Towards Explainable Conservation Monitoring via Attribution and Perturbations
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912669148119040 |
|---|---|
| author | Zhou, Jiayi Aghakishiyeva, Günel Arya, Saagar Dale, Julian Poling, James David Houliston, Holly R. Womble, Jamie N. Larsen, Gregory D. Johnston, David W. Bent, Brinnae |
| author_facet | Zhou, Jiayi Aghakishiyeva, Günel Arya, Saagar Dale, Julian Poling, James David Houliston, Holly R. Womble, Jamie N. Larsen, Gregory D. Johnston, David W. Bent, Brinnae |
| contents | Computer vision can accelerate ecological research and conservation monitoring, yet adoption in ecology lags in part because of a lack of trust in black-box neural-network-based models. We seek to address this challenge by applying post-hoc explanations to provide evidence for predictions and document limitations that are important to field deployment. Using aerial imagery from Glacier Bay National Park, we train a Faster R-CNN to detect pinnipeds (harbor seals) and generate explanations via gradient-based class activation mapping (HiResCAM, LayerCAM), local interpretable model-agnostic explanations (LIME), and perturbation-based explanations. We assess explanations along three axes relevant to field use: (i) localization fidelity: whether high-attribution regions coincide with the animal rather than background context; (ii) faithfulness: whether deletion/insertion tests produce changes in detector confidence; and (iii) diagnostic utility: whether explanations reveal systematic failure modes. Explanations concentrate on seal torsos and contours rather than surrounding ice/rock, and removal of the seals reduces detection confidence, providing model-evidence for true positives. The analysis also uncovers recurrent error sources, including confusion between seals and black ice and rocks. We translate these findings into actionable next steps for model development, including more targeted data curation and augmentation. By pairing object detection with post-hoc explainability, we can move beyond "black-box" predictions toward auditable, decision-supporting tools for conservation monitoring. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_21689 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | On Thin Ice: Towards Explainable Conservation Monitoring via Attribution and Perturbations Zhou, Jiayi Aghakishiyeva, Günel Arya, Saagar Dale, Julian Poling, James David Houliston, Holly R. Womble, Jamie N. Larsen, Gregory D. Johnston, David W. Bent, Brinnae Computer Vision and Pattern Recognition Artificial Intelligence Computer vision can accelerate ecological research and conservation monitoring, yet adoption in ecology lags in part because of a lack of trust in black-box neural-network-based models. We seek to address this challenge by applying post-hoc explanations to provide evidence for predictions and document limitations that are important to field deployment. Using aerial imagery from Glacier Bay National Park, we train a Faster R-CNN to detect pinnipeds (harbor seals) and generate explanations via gradient-based class activation mapping (HiResCAM, LayerCAM), local interpretable model-agnostic explanations (LIME), and perturbation-based explanations. We assess explanations along three axes relevant to field use: (i) localization fidelity: whether high-attribution regions coincide with the animal rather than background context; (ii) faithfulness: whether deletion/insertion tests produce changes in detector confidence; and (iii) diagnostic utility: whether explanations reveal systematic failure modes. Explanations concentrate on seal torsos and contours rather than surrounding ice/rock, and removal of the seals reduces detection confidence, providing model-evidence for true positives. The analysis also uncovers recurrent error sources, including confusion between seals and black ice and rocks. We translate these findings into actionable next steps for model development, including more targeted data curation and augmentation. By pairing object detection with post-hoc explainability, we can move beyond "black-box" predictions toward auditable, decision-supporting tools for conservation monitoring. |
| title | On Thin Ice: Towards Explainable Conservation Monitoring via Attribution and Perturbations |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2510.21689 |