Automated identification of individual animals becomes ecologically useful only when development and evaluation are centered on specific ecological questions, available data realities, and the differing costs of different kinds of errors.
In: 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), IEEE, pp 2090–2100
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
q-bio.PE 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Centering Ecological Goals in Automated Identification of Individual Animals
Automated identification of individual animals becomes ecologically useful only when development and evaluation are centered on specific ecological questions, available data realities, and the differing costs of different kinds of errors.