Abstract
Long-term animal re-identification (ReID) must remain robust to gradual morphological evolution and seasonal appearance shifts. While recent vision-language models provide strong pretrained visual representations, adapting them to longitudinal ecological settings remains challenging, particularly under identity and temporal distribution shifts. We present a parameter-efficient CLIP adaptation framework for animal ReID and introduce a continuous metadata-conditioning mechanism that incorporates numerical attributes directly into the prompt representation during training.
Low-rank visual adaptation, prompt-based supervision, and cross-modal alignment provide the adaptation framework, while the proposed metadata-conditioning strategy constitutes the primary methodological contribution. By preserving the continuous structure of numerical metadata rather than discretizing it into textual categories, the proposed approach enables smooth modulation of the embedding space during training while maintaining a purely visual inference pipeline.
Experiments on a seven-year longitudinal fish dataset and multiple wildlife benchmarks demonstrate improved performance under closed-set, open-set, and time-aware evaluation protocols. The results indicate that continuous metadata conditioning improves robustness to longitudinal appearance variation and temporal distribution shifts, while parameter-efficient adaptation enables a purely visual inference pipeline without requiring metadata at test time.
Code and evaluation splits can be found at: GitHub.
Blogger's Review: This study significantly enhances animal re-identification's adaptability in dynamic environments by introducing a continuous metadata conditioning mechanism. This approach not only optimizes the efficiency of model parameters but also improves the fluidity of visual inference, showcasing its potential applications in ecological monitoring.